4. DISCUSSION
The prevalence of obstructive sleep apnea with concomitant daytime drowsiness is between 3 and 7 percent for adult males and 2 to 5 percent for adult women [1,3]. OSA is defined by recurrent blockage of the upper airway during sleep, which is associated by episodic hypoxia, wakefulness, and sleep fragmentation. Upper airway anatomic anomalies, such as a thicker palate, hypertrophic tonsils, or a thickened pharynx with decreased neural activity, lead to obstructive apneas and hypopneas in OSA. OSA causes long-term damage to the cardiovascular, neurologic, and respiratory systems. According to a meta-analysis, OSA is associated with increased risks of significant adverse cardiac events, coronary heart disease, stroke, and cardiac death. OSA is also related to inflammation, since repeated bouts of hypoxia increase systemic oxidative stress and contribute to the development of a systemic inflammatory state.[4] Recent research has discovered higher exhaled nitric oxide in OSA exhaled air, which contributes to upper airway inflammation.[5] In this regard, OSA has been associated with the alteration of voice resonance and articulation through thickened soft palate, hypertrophic tonsils, or a thickened pharynx and tongue base lymphoproliferation.[6] However, there has been little research in the effect of OSA on voice to yet.
A cross-sectional analysis of data from the national longitudinal research found a relationship between OSA and vocal disorders. People with OSA symptoms exhibited a higher prevalence of voice abnormalities than those without OSA symptoms (6.7 percent vs. 4.7 percent) in a study of 14,794 young adults in the United States.[6] N. Roy et al. also discovered that 28% of OSA patients had vocal abnormalities, which was greater than the general population.[7] However, in these research, vocal abnormalities were assessed by patients’ subjective voice pain, rather than objective voice characteristics or diseases like BVFL.
The mouth cavity, pharynx, and larynx  known as the vocal tract, are structures that influence speech production. Speech impairment is caused by functional or structural abnormalities in these structures. The Bernoulli effect and enhanced pharyngeal dynamic compliance can explain pharyngeal narrowing and thickening in OSA. As a result, it is assumed that increased vocal tract compliance leads to OSA resonance and articulation problems. Many studies have observed vocal alterations following OSA surgical treatment.[8,9] Eun et al., for example, found that uvulopalatopharyngoplasty alters the formant frequencies of vowels, resulting in alterations in resonance after OSA surgery.[8] However, the changes that occur in the larynx in OSA patients have received less attention. Elongated epiglottis and redundancy in this structure can cause collapse and alterations in glottic and supraglottic structures.[10] Furthermore, Krieger et al. proposed that OSA causes repeated glottic spasm and paradoxic glottic narrowing.[11]
OSA patients frequently breathe via their mouths, which produces a reduction in moisture in the inhaled air, leading the vocal cords to dry. Water loss from the sol layer caused by oral breathing increased the viscosity of respiratory epithelium and overlaying mucus, increased tracheal mucus velocity, and reduced mucociliary clearance.[12] Normal human oral breathing for 15 minutes revealed effects that are most likely the result of superficial dehydration on the vocal cords and increased vocal effort.[13] In a research of air inhalation, Hemler et al. discovered that perturbation measurements were significantly higher after inhaling desiccated air than ordinary air.[14] Many research, on the other hand, have observed low phonation threshold pressures in people exposed to hydrated or ”wet” circumstances.[15] In summary, OSA patients have long-term mouth breathing, which causes higher vocal effort due to superficial dehydration of the vocal fold.
As previously stated, there have been few reports on the impact of OSA on voice. For example, Monoson and Fox and Fox et al. found that the relationship between OSA and voice disorder and discovered that 60 to 70% of OSA patients exhibited a combination of phonation, articulation, and resonance abnormalities.[16,17] Based on objective acoustic parameters, Wei et al. recently showed that OSA patients had vibration irregularity, inadequate glottal closure, and greater hoarseness compared to normal individuals.[18]
Unlike prior research that investigated at voice quality, we used a nationwide cohort study to evaluate at the prevalence of BVFL in OSA patients. Multiple studies, including the National Institutes of Health (NIH) epidemiology study, define voice disorder as ”anytime the voice does not work, perform, or sound as it normally should, so that it interferes with communication,” implying that it is defined based on subjective symptoms based on individual judgment.[19] Meanwhile, BVFLs such as vocal nodules, vocal polyps, and Reinke’s edema are pathologic changes in the superficial layer of the lamina propria that otolaryngologists identify with a laryngoscope. These lesions can be caused by voice abuse, misuse, smoking, alcohol consumption, or larygopharyngeal reflux (LPR). However, the influence of OSA as a cause of BVFL has yet to be studied. In this study, the prevalence of benign vocal fold lesions (BVFL) was reported to be 1.79 times greater in the OSA group than in the control group. Those who had OSA surgery were 35% less likely to be diagnosed with BVFL during the study period.
In a subgroup analysis, the HR for BVFL was greater in female OSA patients than in male OSA patients (HR: 1.22[1.10-1.35]). Females are more likely to have voice disorders in general, and females had a greater rate of BVFL in the literature.[20] Female vocal folds contain less hyaluronic acid in the superficial layer of the lamina propria, resulting in a lower absorption ability to endure phonotrauma, according to Butler et al..[21]
Other risk factors for BVFL include being aged from 40 to 60 years (HR: 1.20 [1.09-1.32]), living in a major city (HR: 1.39 [1.23-1.59]), and having a higher socioeconomic status (HR: 1.10[1.01-1.21]). Roy et al. showed that among randomly chosen participants, the age range of 40–59 years appeared to constitute a high-risk age group for the reporting of voice problems.[22] Both of these findings are similar to the findings of Russell et al., who found that teachers over the age of 50 had higher voice difficulty than younger teachers.[23] Hur et al. identified health inequalities among people in the United States with voice difficulties, indicating that racial minorities and those with low income tend to avoid treatment owing to a lack of transportation.[24] It is well recognized in the literature that persons with poor income or insufficient health insurance face greater hurdles accessing medical services.[25] Similarly, the greater prevalence of BVFL in metropolitan regions can be explained.
We used PS matching in our study to reduce selection bias and confounding variables between the two groups. We used PS matching with the following characteristics to achieve a fair comparison between groups: age, gender, residence, socioeconomic status, and underlying diseases. This is the first study to examine the risk of BVFL in OSA patients using large-scale real-world data. Although we produced significant results, there are some limitations that need be addressed in future study. First, despite the fact that this registry-based method resulted in a large number of unselected research participants, information about classic BVFL risk factors such as work environment, psychological factors, personality traits, voice abuse, and smoking was insufficient. Second, in our analysis, surgical treatment of OSA was related with a lower incidence of BVFL. We were unable to compare the impact of continuous positive airway pressure (CPAP), which is more often used for OSA therapy, due to a lack of data in the NHIS-NSC database, as CPAP has only recently become covered by Korean medical insurance. Third, because OSA was only recognized using diagnostic codes, the severity of the condition could not be determined.
According to this observational study, OSA is related with an increased incidence of BVFL. Subgroup analysis revealed that the incidence of BVFL in OSA patients increased with age, female sex, and high socioeconomic status. Our study also found that surgical correction of OSA decreased the incidence of BVFL. Physicians should be aware of the possibility of developing BVFL in OSA patients, which leads to poor voice outcomes.