Change in Vocal Resonance Following Sinonasal Surgery Authors: Meha G Fox, Sarah L Blumhardt, Lauren McCray, K Kelly Gallagher, Julina Ongkasuwan, Tran B Locke Published: 2026 ABSTRACT ABSTRACTObjectivesDelineate objective and clinician‐perceived subjective changes in voice quality following sinonasal surgery.MethodsProspective cohort study of six adults who completed pre‐ and post‐operative voice recordings which were analyzed for jitter, shimmer, noise to harmonic ratio (NHR), cepstral peak prominence (CPP), formant and anti‐formant changes. Consensus auditory‐perceptual evaluation of voice (CAPE‐V) assessment was completed by two independent reviewers. Wilcoxon signed‐rank test was used to analyze the pre‐ and post‐operative voice measures and CAPE‐V scores.ResultsFour males underwent endoscopic sinus surgery, including two with concurrent septoplasty, and two females underwent septoplasty with inferior turbinate reduction. There was a statistically significant difference between pre‐ and post‐operative formants 2 (F2) and 3 (F3) for the nasal consonant/n/. There was no statistically significant difference in pre‐ and post‐operative formants 1–4 (F1–F4) for the nasal sounds /m/, /Ɛ/, and /i/, or in F1 and F4 for the sound /n/. However, the change in mean for F2–4/m/, F4/a/, F1–4/n/, and F2–4/i/ was greater than 50 Hz, which is clinically significant. There was no statistically significant difference in pre‐ and post‐operative jitter, shimmer, CPP, NHR, or CAPE‐V scores.ConclusionPrior studies demonstrate variable changes in voice following sinonasal surgery. Our study revealed a clinically significant difference in formants. Lack of statistical significance is likely due to a small study population. Larger scale studies assessing objective and subjective measures of voice are planned to better define voice changes following sinonasal surgery.Keywords: paranasal sinus, septoplasty, sinus surgery Introduction Voice is an important aspect of communication that impacts social and vocational functioning. Voice is created by glottic vibrations that are modified by the resonance chambers of the pharynx, oral cavity, and nasal cavity. Endoscopic sinus surgery (ESS) and septoplasty alter the shape and volume of the nasal passages; therefore, these procedures may change voice quality due to altered resonance. Few studies have addressed subjective and objective voice outcomes following sinonasal surgery, and the existing literature shows a variable impact [1]. Prior work has described variable changes in objective and subjective measures of voice. Some studies have examined acoustic parameters such as fundamental frequency (F0), jitter, shimmer [2], harmonics‐to‐noise ratio (HNR) [3], noise‐to‐harmonic ration (NHR) and found no statistically significant changes following sinonasal surgery [2, 3, 4, 5, 6]. A few studies have shown increased objective measures of nasality after surgery [7, 8, 9], but this effect diminished within 6 months [9]. Few studies have assessed the impact of surgery on formants and anti‐formants (also known as antiresonances) which are more representative of sinonasal contribution to voice; Xiao et al. [1] commented that there were insufficient studies regarding formant changes to perform statistical analysis. Previously assessed subjective measures of voice changes include patient‐reported outcomes including the Voice Handicap Index (VHI) [10, 11], voice‐related quality of life [3], validated questions of nasality [12], non‐validated questionnaires [13, 14, 15, 16], and speech language pathologists’ (SLP) evaluation of the grade‐roughness‐breathiness‐asthenia‐strain (GRBAS) scale [5, 9, 12, 17]. No prior studies have assessed consensus auditory‐perceptual evaluation of voice (CAPE‐V), which is inclusive of GRBAS, and arguably more comprehensive. Our prior work on the relationship between subjective voice quality and sinonasal disease severity showed that when stratified into mild, moderate, and severe sinonasal disease, those with moderate disease had 77% lower odds of having an abnormal VHI‐10 score compared to those with severe disease [18]. Since the severity of sinonasal disease correlated to a subjective decline in vocal quality, decreasing disease burden through sinonasal surgery has the potential to improve vocal quality. The present study compares objective and clinician‐perceived subjective measures of vocal functioning, particularly resonance, before and after sinonasal surgery. Methods This study was approved by the institutional review board at Baylor College of Medicine. Patients were recruited from an academic tertiary care rhinology practice. Inclusion criteria were English‐speaking patients over the age of 18 undergoing primary sinonasal surgery. Exclusion criteria were (1) non‐English speaking patient, (2) age under 18, (3) prior history of sinonasal surgery, (4) pre‐existing laryngeal pathology confirmed by laryngoscopy at the time of recruitment, (5) focal or systemic disease that could possibly alter voice production. Patients completed a VHI‐10 and underwent voice recordings prior to surgery and at least 6 weeks after surgery. This time point was chosen to ensure adequate post‐operative healing. Voice recordings were performed by licensed SLPs using the Computerized Speech Lab (CSL) (Version 4.0.0) with the 4500 external module from Pentax Medical (Montvale, NJ). Each patient was instructed to say/a/: (1) at a comfortable pitch and loudness, (2) with a loud voice, (3) a quiet voice, (4) a high pitch, and (5) a low pitch. Each patient was also recorded reading the rainbow passage (Appendix S1), which is phonetically balanced and widely used for objective voice analysis. Analysis was then completed using the multidimensional voice program (MDVP), analysis of dysphonia in speech (ADSV) programs, and Praat, an acoustic analysis program. The modal/a/recordings were ultimately chosen to be analyzed for measures of jitter, shimmer, and NHR based on similar studies [3, 8]. Cepstral peak performance (CPP) was obtained using the rainbow passage recordings. Formant (F1–F4) and anti‐formant (A1–A3) analysis was completed using a custom Praat script created for this project by Ian Howell, DMA (Appendix S2). The word “many” from the rainbow passage was chosen for formant and anti‐formant analysis given its phonetic features. A spectrogram of each individual phoneme, /m//Ɛ//n//i/, was created within Praat for analysis. The custom Praat script was then used to obtain F1–4 data for the vowels /Ɛ/ and /i/. Anti‐formants are unique to certain consonants, therefore F1–4 and A1–A3 data were obtained for /m/ and /n/. CAPE‐V assessment of the recordings was completed by two independent reviewers (J.O. and S.B.), a senior laryngologist and SLP. Wilcoxon signed‐rank test was used to analyze the pre‐ and post‐operative acoustic measures and CAPE‐V scores. A statistician completed statistical analyses using R (Version 4.0.4) and RStudio (Version 3.0.386). Results Four males underwent ESS, two of whom had concurrent septoplasty, and two females underwent septoplasty with inferior turbinate reduction (Table 1). The average age of participants at the time of surgery was 40 (range: 18–70). Post‐operative recordings were done an average of 95.83 days post‐operatively. This was based on post‐operative healing and patient availability for recordings. Patients were evaluated endoscopically to ensure healing prior to post‐operative recordings. No adverse events were encountered over the course of the study period. Table 1Demographics and surgery type of sinonasal surgery patients.PatientSexAge at surgery (year)Type of surgery1M57ESS (bilateral maxillary antrostomy, total ethmoidectomy, sphenoidotomy, frontal sinusotomy with preservation of middle turbinate).2M18Septoplasty, ESS (bilateral maxillary antrostomy, total ethmoidectomy, sphenoidotomy, frontal sinusotomy with preservation of middle turbinate).3M59ESS (bilateral maxillary antrostomy, total ethmoidectomy, sphenoidotomy with preservation of middle turbinate).4F18Septoplasty, inferior turbinate reduction.5F19Septoplasty, inferior turbinate reduction.6M70Septoplasty, ESS (bilateral maxillary antrostomy, total ethmoidectomy, sphenoidotomy, frontal sinusotomy with preservation of middle turbinate).Abbreviation: ESS, endoscopic sinus surgery. Demographics and surgery type of sinonasal surgery patients. Abbreviation: ESS, endoscopic sinus surgery. There was a statistically significant difference between pre‐ and post‐operative formants 2 (F2) and 3 (F3) for the nasal consonant/n/ (Table 2). There was no statistically significant difference in pre‐ and post‐operative formants 1–4 (F1–F4) for the nasal sounds /m/, /Ɛ/, and /i/, in F1 and F4 for the sound /n/, or in anti‐formants (A1–A3) for consonants /m/ and /n/ (Table 2). However, the change in mean for F2–4/m/, F4/Ɛ/, F1–4/n/, and F2–4/i/was greater than 50 Hz (Table 2), which is clinically significant [19, 20]. There was no statistically significant difference in pre‐ and post‐operative acoustic measures jitter, shimmer, NHR, and CPP, or CAPE‐V scores (Table 3). Table 2Mean difference in formant and anti‐formant scores of sinonasal surgery patients (Hz, n = 6).Pre‐operative mean formant/anti‐formant score (SD)Post‐operative mean formant/anti‐formant score (SD)Mean differencea p value /m/ F1 356.08 (84.55)372.92 (41.21)−16.840.56 F2 1457.51 (275.55)1764.79 (262.39)−307.280.22 F3 2552.61 (273.23)2765.20 (95.11)−212.590.22 F4 3887.54 (210.26)4029.13 (230.01)−141.590.31 A1 881.92 (51.60)841.47 (49.88)40.450.28 A2 1694.83 (292.21)1756.63 (327.25)−61.790.84 A3 2873.28 (99.57)2712.11 (304.14)161.170.56 /a/ F1 482.72 (113.96)476.18 (113.23)6.540.31 F2 1882.10 (191.86)1921.31 (225.91)−39.210.69 F3 2611.49 (291.67)2569.78 (357.22)41.710.16 F4 3848.35 (398.34)3679.98 (401.33)168.370.56 /n/ F1 356.71 (62.04)420.33 (50.14)−63.620.06 F2 1429.42 (338.20)2057.71 (196.85)−628.290.03 F3 2482.80 (275.18)2900.73 (362.36)−417.930.03 F4 3721.83 (433.01)4175.89 (522.46)−454.060.22 A1 811.89 (119.92)885.97 (34.36)−74.080.44 A2 1486.51 (303.13)1607.93 (202.95)−121.410.56 A3 2826.16 (110.36)2639.52 (321.72)186.640.31 /i/ F1 414.38 (99.84)393.24 (43.33)21.140.84 F2 1942.89 (134.34)2253.16 (316.48)−310.270.16 F3 2663.43(261.66)2761.14 (291.11)−97.710.69 F4 3738.37 (557.08)3947.31 (474.80)−208.940.56 a Mean difference: pre‐post. Mean difference in formant and anti‐formant scores of sinonasal surgery patients (Hz, n = 6). Mean difference: pre‐post. Table 3Mean difference in consensus auditory‐perceptual evaluation of voice (CAPE‐V) scores of sinonasal surgery patients (n = 6).Pre‐operative mean score (SD)Post‐operative mean score (SD)Mean differencea p valueSeverity20.83 (18.82)17.50 (9.87)3.331.00Roughness20.83 (11.58)15.83 (14.63)50.34Breathiness5 (7.75)5 (5.85)01.00Strain10.00 (11.40)5.83 (5.85)4.170.85Pitch0.00 (0.00)0.00 (0.00)0—Loudness2.50 (6.12)3.33 (5.16)−0.831.00Resonance—hypernasal, hyponasal, cul‐de‐sac, or normal2.67 (2.36)1.17 (1.33)1.500.10 a Mean difference: pre‐post. Mean difference in consensus auditory‐perceptual evaluation of voice (CAPE‐V) scores of sinonasal surgery patients (n = 6). Mean difference: pre‐post. Discussion Within the source‐filter model of speech production, the nasal cavity functions as one of the vocal tract filters [21]. A change in the size and contents of the nasal cavity after sinonasal surgery could therefore impact voice quality, particularly resonance. Resonance can be measured in a variety of ways including, but not limited to, nasalance (nasal acoustic energy as a percentage of total acoustic energy from the mouth and the nose) [1], flexible endoscopy, spectrogram analysis, and articulation assessment. A recent systematic review and meta‐analysis found statistically significant changes in nasalance at 1 month after surgery (ESS, septoplasty, rhinoplasty, turbinate reduction), but this difference did not persist at 6 months after surgery. Other objective measures demonstrated no statistically significant difference. The authors do comment that spectrogram formant analysis was not completed in enough studies of patients undergoing ESS to perform statistical analysis [1]. Data regarding voice changes following sinonasal surgery remain limited to small prospective series focused on features that occur at larynx including jitter, shimmer, HNR, and NHR along with subjective patient reported outcomes (PROs) or clinician graded scales [1]. We expected that speech characteristics of the voice filter or resonators, in particular vowel formants and nasal formants and anti‐formants, are more likely to change after sinonasal surgery. Formants are amplified frequency ranges of vowels seen on a spectrogram. Each vowel, produced with an open oral cavity and vibrating vocal folds, has unique formant characteristics due to the specific configurations of the tongue, lips, and jaw. Three to four formants are expected to be seen on spectrogram depending on the vowel. Nasal sounds differ from vowels in their production and formant characteristics. Nasal sounds are created by allowing a degree of airflow into the nasal cavity while the oral cavity is closed or nearly closed, and the vocal folds are vibrating. In this configuration, nasal membranes absorb much of the sound energy. Therefore, on spectrogram, the amplitude of F1–F3 in nasal consonants is lower than the amplitude of vowels. Additionally, the closed oral cavity serves as a “side” resonator and lends to anti‐formants, which are identifiable on spectrogram [21]. To our knowledge, only four studies have assessed formant frequencies before and after septoplasty [4, 6, 10, 12], and only two addressed formant changes following ESS [14, 22]. These studies showed no statistically significant difference in F1–4 [10, 12], F3–4 for vowels /i, e, a, o, u/, or F1–F4 for nasal consonants /m, n/ and nasalized vowel /i/ [4, 6] following septoplasty. Similarly, there was no significant change in formants following ESS [14], but Hong et al. [22] did demonstrate a slight decrease in F1 3 weeks after ESS for nasal polyposis. In our study, we found a statistically significant difference in pre‐ and post‐operative formants (F2–3) for the nasal consonant /n/. We also discovered a change in mean greater than 50 Hz for several formants which is clinically significant [19, 20]. Some of these changes may not have reached statistical significance due to small sample size. Prior studies have assessed subjective changes in voice following sinonasal surgery with heterogeneous methods. Patient perceptions have been assess using validated PROs such as voice‐related quality of life (VRQOL) [3] or VHI [10, 11], using a visual analog scale (VAS) for patient satisfaction with voice [5], using unvalidated questionnaires [15, 16], or simply asking a patient if they perceived a voice change [4, 6, 14, 17]. Provider perception has been evaluated using the GRBAS scale completed by SLPs [5, 9, 12, 17]. These studies demonstrated variable subjective outcomes. Of the four studies assessing GRBAS, one showed no significant difference [5], one showed a difference of less than two points [9], and two showed a significant difference following ESS and septoplasty [12, 17]. VHI decreased following septoplasty in one study [10], while VRQOL had no change [3]. No prior studies have included CAPE‐V to assess provider perception of voice changes following sinonasal surgery. The CAPE‐V was established for providers to have a common method for evaluating severity of auditory‐perceptual attributes of voice. It includes components of the GRBAS. In our limited sample, we found no change in CAPE‐V scores following sinonasal surgeries when completed by a senior laryngologist (J.O.) and an experienced SLP (S.B.). Limitations Two limitations of this study were small sample size and heterogeneous surgeries. While recruitment occurred in a tertiary care rhinology clinic, volume was limited by the COVID‐19 pandemic and duration of the project funding. Additionally, we had strict inclusion criteria that eliminated potential candidates. Specifically, English‐speaking patients who had no prior sinonasal surgeries were limiting criteria. In our tertiary care academic practice in one of the nation's largest and most diverse cities, we see many non‐English‐speaking patients who have had prior surgical interventions. Nasalance is another measure of nasal contribution to voice. We were unable to capture this measure due to lack of equipment; however, we have since acquired funding for the equipment and are currently studying this in additional subjects. Conclusion Prior studies demonstrate variable changes in voice following sinonasal surgery. Our study revealed a clinically significant difference in some but not all formants. Lack of statistical significance is likely due to a small study population. Larger scale studies assessing objective and subjective measures of voice are needed to better define voice changes following sinonasal surgery. Author Contributions Meha G. Fox recruited patients, assisted with data collection, and authored the manuscript. Sylvia Adu‐Gyamfi completed statistical analysis and contributed to the methods and results sections of the manuscript. Sarah Blumhardt completed voice recordings, completed the CAPE‐V assessment, assisted with manuscript development and editing. Lauren McCray assisted with data collection, organization, and manuscript development and editing. K. Kelly Gallagher assisted with manuscript editing. Julina Ongkasuwan assisted with study design, completed the CAPE‐V assessment, and manuscript editing. Tran B. Locke designed the study, collected data, analyzed data, and edited the manuscript. Ethics Statement This study was approved by the institutional review board at Baylor College of Medicine. Patients were recruited from an academic tertiary care rhinology practice. Conflicts of Interest Meha G. Fox served on an advisory board for Sanofi Regeneron Pharmaceuticals and as a consultant for Aerin Medical. The remaining authors declare that there is no conflict of interest. This work was presented as a poster presentation at the Annual Meeting of the American Rhinologic Society in 10/2023 in Nashville, TN, USA. Supporting information Supporting File 1 Supporting File 2 Acknowledgments A special thanks is also given to Ian Howell, DMA, from Embodied Music Lab in Ann Arbor, Michigan for his creation of the Praat script used for formant and anti‐formant analysis, and to Sylvia Adu‐Gyamfi, MPH, for completing the statistical analysis during her time in the Baylor College of Medicine Center of Excellence in Health Equity, Training and Research. This work was supported by the American Rhinologic Society Short‐Term Mentorship Program Research Grant and mentor David Francis MD. Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request. References Associated Data Supplementary Materials Supporting File 1 Supporting File 2 Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request.