To our best knowledge, although snoring is the most common symptom in OSA, no international consensus exists regarding its objective definition based on PSG results. Our study suggested that the snoring episode index—the first concept we developed as an indicator related to snoring—highly correlated with the AHI. This showed the potential of utilizing the snoring episode index as a definition of snoring related to the severity of OSA. We conducted this study to identify an index that highly correlated with sleep quality or OSA severity by using snoring-related data that can be easily obtained using PSG, and found that the snoring episode index met this purpose.
Many attempts have been made to define snoring both objectively and easily, but most of them have had some limitations. Initially, attempts were made to assess the severity of snoring by administering questionnaires to snorers or their sleep partners. In one such questionnaire study on snorers and their spouses, subjective factors inevitably played a role, as men underestimated their own snoring and women overestimated their spouse’s snoring. Another study aimed to define snoring by using a VAS. The 10 snoring samples were evaluated by 53 evaluators on a 50-step VAS. The results of that study showed high consistency and concordance, but when the peak level was adjusted to exclude the effect of different volumes of each snoring, the concordance was found to be low. Efforts have also been made to objectively define the intensity of snoring. However, owing to the lack of an international consensus on sound intensity for the definition of snoring, different criteria were applied depending on the investigators. For example, 40 dB or 50 dB was used as a loudness threshold to define snoring.[12,20] Efforts have also been made to define snoring according to its frequency. In a study examining the relationship between the frequency of snoring per hour and sleep-related parameters in 74 people, the frequency of snoring per hour and the AHI showed a weak positive correlation. However, the criteria used for defining the frequency of snoring was unclear. In another study, snoring frequency was defined as the percentage of inspiratory breaths during sleep with sound loudness peaks ≥ 40 dB, and snoring intensity was defined as the mean peak inspiratory sound loudness, which showed a moderate positive correlation with the AHI. Nevertheless, in a literature search, we could not find any studies using a definition of snoring that showed as strong a correlation with the AHI as did the newly suggested snoring episode index.
Before using the snoring episode, which has been used conventionally in our PSG laboratory, as a new definition of snoring, we needed to validate the new concept of a snoring episode. In this study, we finally defined 3 consecutive snoring events as one snoring episode and produced a snoring episode index by dividing the total number of snoring episodes by the total sleep time, proving that the snoring episode index very strongly correlated with the AHI. We defined a single snoring event as a >200-microbar nasal airflow pressure recorded using a nasal pressure transducer, and we conducted a study to optimize the number of snoring events included in a single snoring episode. The number of consecutive snoring events included in one snoring episode was set to 1, 2, 3, 4, and 5; thereafter, the entire PSG recording was reread to calculate the snoring episode index in each case, and the number of consecutive snoring events with the best correlation with the AHI was set to define the snoring episode. When 2 or 3 consecutive snoring events were included in one snoring episode, the correlation between the snoring episode index and the AHI was the highest. Therefore, we determined that the conventional use of 3 consecutive snoring events for the definition of one snoring episode was reasonable.
In this study, four parameters related to snoring were used, namely, the snoring episode index, snoring percentage, average snoring episode duration, and longest snoring episode duration, and their correlations with several sleep-related factors were analyzed. The BMI, which is closely related to snoring, showed the highest positive correlation with the snoring episode index among the four parameters. The ESS, a questionnaire reflecting daytime sleepiness (one of the symptoms that indirectly reflects the quality of sleep), also showed the highest positive correlation with the snoring episode index. Meanwhile, the PSQI, a comprehensive survey on sleep quality, showed a negative correlation with the snoring episode index, indicating that it was difficult to accurately predict sleep quality on the basis of snoring. Other major PSG parameters, such as sleep latency, sleep efficiency, rapid eye movement sleep percentage, and supine sleep time percentage, did not show high correlations with snoring parameters.
In one study in the United States, 93% of women and 82% of men with moderate to severe OSA were undiagnosed, as were 98% of women and 90% of men with mild OSA. To overcome this issue of underdiagnosis, timely prescreening of OSA is important. If the concept of the snoring episode index presented in this study is included in contactless home sleep tests, the performance of prescreening based on smartphone recording applications may be enhanced. Nevertheless, the snoring episode index proposed in this study has a limitation. The snoring episode index was calculated on the basis of snoring assessed using an airflow pressure sensor. Therefore, it should be validated when snoring is assessed using a smartphone’s sound recording system. In the future, we plan to conduct a prospective study to validate the snoring episode index by utilizing snoring data obtained using sound recording applications on smartphones or artificial intelligence speakers.
In conclusion, we suggest a new index for diagnosing snoring, i.e., the snoring episode index, and show that this index highly correlates with the AHI. Although snoring can be defined from various perspectives, the strong correlation between the snoring episode index and OSA severity has a clinical implication given that this index can be used for the prescreening of OSA. Future validation studies using various sound recording systems are required as they will extend the utility of the snoring episode index to home sleep screening applications and devices.