2.1 Participants
The cross-sectional study was approved by the Ethics Committee of Air Force General Hospital, China, and all work was carried out in accordance with The Code of Ethics of the World Medical Association (Declaration of Helsinki). It was conducted in an establishment’s male dormitory compound under full-close management in Beijing during June to August, 2019.
The calculation of sample size was performed according to the method reported by Dupont et al.[15] Based on previous publications[10-13], the Odds Ratio (OR) of OSA in periodontitis is approximately 2. So, we set twofold increase in the prevalence of periodontitis in OSA group compared with control group as clinical important, and the sample size of 90 reached study power >90% with a two-sided significance test α = 0.05. Due to the relevance of smoking and periodontitis[16], the inclusion criteria were: 1) Age > 18 years old; 2) male; 3) self-reported without alcohol abuse; 4) self-reported never smoking or having quitted smoking for more than 6 months. Considering the complication of the study, we initially recruited 135 participants from all residents at the dormitory compound using simple random sampling method.
All participants signed informed consent after elucidating study details. The exclusion criteria are set as follows: 1) lack complete medical records; 2) have systematic diseases that could affect OSA and periodontitis; 3) have specific use of medication that could affect OSA and periodontitis; 4) currently diagnosed and/or under treatment of OSA or periodontitis; 5) with genetic susceptibility; 6) have special diet preference. Among the originally included 135 participants, 119 underwent medical check within one year and provided with complete medical history. The elimination reasons were: 1) Blood-system diseases (n = 1); 2) Uncontrolled severe hypertension, SBP/DBP > 180/110mmHg (n = 1); 3) Diabetes mellitus (n = 1); 4) Medication history of phenytoin, calcium channel blocker, cyclosporin (n = 2); 5) Currently under treatment of OSA (n = 1) or periodontitis (n = 2).
Before further examinations, participants were measured body mass index (BMI), completed a questionnaire related to sleep factors and life habits and passed the pre- polysomnography (PSG) check. Two participants voluntarily quitted the periodontal examination and the following nasal airway resistance (NAR) assessment, and one failed to record PSG data due to connect failure, leaving a total of 93 completed the whole study. (Fig. 1)
All participants were then divided into OSA group and non-OSA group according to the PSG diagnosis, with a final total 19 in the OSA group and 74 in the non-OSA group. Additional participants demographics are presented in Table 1. Almost all participants could be considered normal range (59.1% had BMI from 18.5 ~ 24.9 kg/m2) or pre-obese (39.8% had BMI from 25 ~ 29.9 kg/m2) according to the WHO criteria for obesity.[17] A higher proportion of pre-obese participants were discovered in the OSA group, and t-test also indicated significantly higher AHI (p < 0.001) than the non-OSA group. Other demographic characteristics including age (p = 0.505) and alcohol use were basically matched between the two groups. (Table 1)
2.2 Diagnosis of OSA
All participant underwent overnight PSG with portable in-lab PSG devices (SOMNOscreenTM plus, SOMNOmedics GmbH, Randersacker, Germany) in the dormitory compound.
Before the PSG, each participant had heart rate and blood pressure recorded. Accordingly, electrocardiogram, central electroencephalogram, bilateral electrooculograms, submental electromyogram, bilateral anterior tibialis electromyogram was taken. Chest and abdominal movement were traced by respiratory effort bands. Body position recorded by pressure sensors. Oronasal airflow was measured by oronasal thermal sensors. And oxygen saturation (SaO2) was measured by a pulse oximeter. In the morning, heart rate and blood pressure were recorded again within one hour after the rising.
Sleep stages, respiratory events, arousals and body movements were manually scored by one experienced sleep technician according to the manual of the American Academy of Sleep Medicine (AASM).[18] To score a respiratory event as an apnea, the followings should be met: 1) There was a drop in the peak signal excursion by ≥ 90% of pre-event baseline using the oronasal thermal sensor. 2) The duration of the ≥ 90% drop in sensor signal was ≥10 seconds. A hypopnea was scored when: 1) The peak signal excursions dropped by ≥ 30% of pre-event baseline using nasal pressure sensor. 2) The duration of the ≥ 30% drop in signal excursions was ≥ 10 seconds. 3) There was ≥ 3% oxygen desaturation from pre-event baseline or the event is associated with an arousal. If a portion of a respiratory event that would meet both the criteria for a hypopnea and an apnea, the entire event should be scored as an apnea. The apnea–hypopnea index (AHI) is defined as apneas plus hypopneas per hour of sleep. In adults, OSA is diagnosed when AHI score is ≥ 5. OSA severity classification is based on AHI scores, as mild OSA is diagnosed as an AHI of at least 5 to 15, moderate OSA as AHI >15 to 30, and severe OSA as AHI >30.[18] The lowest SaO2 and oxygen desaturation index (ODI) were also recorded as sensitive markers of intermittent hypoxia.[19] ODI is the average number of desaturation episodes per hour, where desaturation is defined as a decrease in the mean SaO2 of ≥4% that lasts for at least 10 seconds over the last 120 seconds.[18]
2.3 Periodontal Examination
The periodontal examination was practiced for each participant by one experienced dentist in the same day immediately before PSG. The examiner was full blinded to the participants’ correlative clinical data in the examination phase and masked to the patient’s OSA conditions in the data collection phase.
Based on these available data and experience from clinical practice[20], the following “Ramfjord Teeth” were examined to indicate the full-mouth periodontal conditions: #3 (maxillary right first molar), #9 (maxillary left central incisor), #12 (maxillary left first bicuspid), #19 (mandibular left first molar), #25 (mandibular right central incisor), #28 (mandibular right first bicuspid).[21] If a Ramfjord tooth was missing, a substitute tooth (#2, #8, #13, #18, #24, #29) was selected as Fleiss et al suggested.[22]
Periodontal parameters including: 1) probing depth (PD), 2) clinical attachment level (CAL), 3) bleeding on probing (BOP) were recorded of six points (mesio-buccal, mid- buccal, disto-buccal, disto-lingual, mid-lingual, mesio-lingual) each tooth in special designed periodontal charts. All measurements were obtained using a Williams style periodontal probe (Hu-Friedy, Inc., Chicago, IL, USA) and lengths were rounded up or down to the nearest millimeter. CAL is determined as follows: 1) If the gingival margin is located on the anatomic crown, CAL is PD minors the distance from the gingival margin to the cemento-enamel junction (CEJ). 2) If gingival margin coincides with the CEJ, CAL equals to PD. 3) If the gingival margin is located apical to the CEJ, CAL is PD plus the distance between the CEJ and the gingival margin. BOP was recorded as either present or absent within 30 seconds of probing.[23]
In diagnosis of periodontitis, we used the clinical case definitions proposed by the Centers for Disease Control and Prevention (CDC), in collaboration with the American Academy of Periodontology (AAP) working group for use in population-based surveillance of periodontitis. Periodontitis is defined as two or more interproximal sites with CAL ≥ 3mm (not on same tooth) and two or more sites interproximal sites with PD ≥ 4mm (not on same tooth) or one site with PD ≥ 5mm. And periodontitis severity classification is defined as: 1) severe periodontitis, when two or more interproximal sites with CAL ≥ 6 mm (not on same tooth) and one or more interproximal site(s) with PD ≥ 5mm; 2) moderate periodontitis, when two or more interproximal sites with CAL ≥ 4 mm (not on same tooth) or two or more interproximal sites with PD ≥ 5 mm (not on same tooth); 3) mild periodontitis, when the patient could not be diagnosed as neither moderate nor severe periodontitis.[24]
To evaluate the severity of periodontitis more accurately, the mean PD, mean CAL and percentage of BOP sites were calculated for each participant.
2.4 Nasal Airway Resistance
It’s suggested that OSA patients may also have a tendency of mouth breathing[25, 26], and a former cross-sectional study concluded a possible association between mouth breathing and gingivitis in children.[27] Then we attempted to record the mouth breathing rate of the study sample. However, surveying mouth breathing by questionnaires could be greatly affected by self-knowledge biases. So, we used NAR as a proxy, as NAR was proved to be an accurate predictor for mouth breathing and subjects with normal nasal airway resistance breathe almost completely through the nasal airway during sleep.[28]
To objectively assess the NAR, both active anterior rhinomanometry (Rhinomanometer NR6; GM Instruments Ltd., Irvine, UK) and rhinospirometry (Rhinospirometer NV2; GM Instruments Ltd., Irvine, UK) were performed on each participant after the periodontal examination and before the PSG. NAR was indicated by recording both nasal pressure and inspiratory nasal airflow to reflect the resistant status of the nasal airway when breathing.
The measuring procedures were conducted in accordance with the guidance of the International Committee on Rhinomanometry Standards.[29] For participants, measurements were obtained in a sitting position, after an adaptation time of 20-30 minutes. Each participant was guided to gently clear the nostrils by blowing, and a cotton swab was used to clear the left secretions.
Rhinomanometry was conducted at first. The pressure nozzle attached by an elastic and fitting plug was placed into one nostril and sealed with adhesive tape. Then, a respiratory face mask with a flowmeter was used, thus providing with a transnasal pressure gradient between two sides. The values of pressure gradient (∆P) and transnasal airflow (V) were monitored by one experienced rhinomanometry technician. Participants were instructed to breathe normally until at least four consecutive inspirations were recorded. The resistance value of the other side of nasal airway was obtained afterwards using the same method.
Rhinospirometry was conducted immediately after the rhinomanometry. The participant was instructed to place the nasal cannulae of the rhinospirometer in seal with the alar rim without deforming the nostril, so as to avoid any leakage. Then, the participant was told to inhale through the nose and exhale through the mouth in normal breathing rate. The volume of inspired air for each nasal passage was recorded simultaneously in a continuous 20 seconds. The measurements were repeated on two further occasions, and mean volumes were calculated.
Before analyzation, the data was examined to eliminate false breaths. Rhinomanometry measurement was expressed by the formula R = ∆p/V, where we picked measurements at ∆p = 150 Pa according to the recommendation of Standardization Committee on Objective Assessment of the Nasal Airway.[29] Rhinospirometry measurement was presented by the mean volume of left and right nasal passage in the 3 successive measurements.
2.5 Statistical Analysis
All statistical analysis was conducted using IBM SPSS Statistics for MacOS, version 26.0 (IBM Corp., Armonk, N.Y., USA), and a two-sided p-value of <0.05 was considered to be statistically significant. A normality test was practiced on all variances we collected, showing that AHI, PD, BOP, and CAL was basically in line with normal distribution.
To investigate the association between OSA and periodontitis, we set OSA as the risk factor and periodontitis as the outcome. Due to the low prevalence of OSA in our study and the lack of moderate/severe OSA patients, we simplified the OSA classification into two dichotomous variables (Yes/No). Then, periodontal parameters including PD, BOP and CAL were firstly tested between OSA group and Non-OSA group. The differences between the mean values of PD, BOP and CAL were evaluated using the t-test. Then, the χ² test was applied to measure the difference in prevalence of periodontitis between variables. And a logistic regression model was used to practice multivariate analysis of periodontitis and exclude confounding factors such as age, BMI and alcohol use.
To identify the possible association between the severity of OSA and that of periodontitis, we took AHI to estimate the severity of OSA according to the recommendation of AASM.[18] Spearman rank correlations were calculated for ordinal variables, and Pearson correlation analysis were used in continuous variances in normal distribution.
Then, to test the possible mechanism that periodontitis was with increased prevalence and severity in OSA patients, we set a binary logistic regression model with five independent variables to calculate ORs and the 95% confidence interval (CI) predicting periodontitis,[30] where AHI, ODI, lowest SaO2, NAR (indicated by rhinomanometry and rhinospirometry measurements) were included as covariates.