Mouth puffing, a derivation of mouth breathing, may possibly be a factor leading to OSA. Since the methodology of measuring mouth puffing is straightforward and easily incorporated into an analytic computer system, the mouth puffing phenomenon may be applied to future physiological research on sleep-related illness. In our study, several insights were revealed: (1) when mouth-taped, some patients with OSA show the symptom of mouth puffing, which can be detected with a MPD, and (2) the relationship between MPSs and both AHI and oxygen-related variables are shown by PSG data.
We have shown that the MP phenomenon does exist, that the MPD is able to identify the MP phenomenon, and that the MP signals can be marked by our algorithm. To verify the efficacy of the MPD, we found a high consistency between the physical observation and the algorithm (in laboratory: correlation = 0.990, p < 0.001), which demonstrated that our algorithm is effective in marking the MPSs. Though MPSs observed from the video recording and MPSs marked by the algorithm showed high correlation, the algorithm showed several kinds of aberrant signals like turning over, coughing, and tooth-grinding, which can be observed from the video recording. In order to identify the MPSs more accurately and to obliterate irrelevant signals, it will be necessary to modify the algorithm. Using a half-minute as the reporting unit, the waves were easily disrupted and the error rate was high. Using a minute as the reporting unit, it was possible to get 10–20 waves at a time, and the error rate was found to be an acceptable 5%. Using the MPSs recorded per minute allowed for correlations among sleep stages, AHI scores, and blood oxygen levels. In our experience, using a minute as the reporting interval is a feasible way of showing the data.
Mouth taping can prevent patients with OSA from inhaling with the mouth but cannot prevent patients from exhaling with the mouth. Many past studies have researched the relationship between oral/nasal breathing and OSA. The relationship between the mouth puffing phenomenon and OSA has now been evaluated for the first time in this study. Similar findings have been reported in other studies that patients who have a higher percentage of oral and oro-nasal breathing periods have more serious OSA and lower SpO2 than common snorers or healthy subjects [10, 12, 13, 15, 25]. However, our findings also indicate that mouth breathing should be divided into two categories, IMP and CMP, defined on a minute by minute basis. Differences among the four MPSs were further investigated. Examination shows that the IMP ratio is positively correlated with the ODI/T90 and negatively correlated with the mean SpO2, when compared with the other MPSs. Participants when completely breathing with nostrils (NMP) tend to have more stable SpO2 during sleep [12, 13, 15]. Though a regular breathing, CMP is positively correlated with lower SpO2 than NMP is, which indicates that CMP is a worse breathing pattern than NMP is.
Oral breathing is a common phenomenon of patients with OSA patients during sleep, happening more frequently right before and after events of apnea and hypopnea. An event of apnea or hypopnea is usually accompanied by a deep and long oral breathe, most likely because the patient tries to make up for oxygen depletion. Oral breathing, accompanied with other factors, causes events of apnea and hypopnea, which in turn causes oral breathing. It is a vicious cycle that oral breathing and events of apnea and hypopnea reinforce each other [15]. Past studies have shown that open mouth breathing tends to cause airway collapse. Mouth breathing patients with OSA tend to have more serious OSA and worse oximetric variables. Also, mouth breathing is associated with more serious and more prevalent lateral pharyngeal wall collapse and tongue base collapse [13].
We theorize that the mouth puffing phenomenon may be an indicator of OSA and may be useful in the diagnosis of OSA. In past studies, a face mask seal has commonly been used to detect airway flow to determine when nasal breathing or oral breathing is present. In the current study, we used the mouth puffing phenomenon to determine when nasal breathing or oral breathing was present and found the same relationship between oral breathing and OSA as other studies have. Similar findings have been reported in other studies that patients who have a higher percentage of oral and oro-nasal breathing periods have more serious OSA and lower SpO2 than common snorers or healthy subjects [11,12,13, 15].
The pathophysiological mechanism is that oral breathing ensues from upper airway resistance, which includes allergic nostril obstruction and constriction of the upper airway caused by bad habitual breathing habits [11, 12]. In our study, patient who used mount breathing tended to have a higher standard deviation of SpO2, a lower mean SpO2, and a lower T90 during sleep. A plausible explanation for the phenomenon is that when the patient breathes with the mouth, the ODI drops or fluctuates, which engenders a higher standard deviation of SpO2, a lower mean SpO2, and a higher T90 during sleep. After a period of apnea a big mouth breathe often occurs as if the patient is trying to catch up after being deprived of oxygen [10, 26].
In supplement materials, we also show that patients with OSA have worse AHI and ODI and higher percentage of IMP during stage REM sleep. In our study, three of ten patients with OSAs had higher AHI, ODI, and IMP during NREM sleep. We suggest that data obtained during stage REM sleep and during NREM be analyzed separately in future studies as we do in this study to avoid data from being skewed. Generally, people breathe more regularly during the NREM sleep and more irregularly during stage REM sleep. People breathe more irregularly during the REM sleep because muscles of pharynx slacken, are slow to inhibit apnea and hypopnea, and cause apnea and hypopnea to occur more frequently and longer. It is found that the geniogiossus muscle becomes very inactive during stage REM sleep, causes the tongue to slide back, causes the airway to be obstructed, and induces apnea and hypopnea. Prior studies have found that 50% of patients with OSA patients belong to the NREM-AHI group [28].
There are several limitations to our study. First, there were a low number of subjects recruited in our study. Second, the data were obtained at the clinic and laboratory, not the patient’s usual sleeping location and time, which may affect the data obtained. Third, this study used a cross-sectional design that suggests only correlations between MPSs and OSA and cannot infer causality.