Follow‐up of patients with mild COVID‐19 using subjective, acoustic, and aerodynamic measurements

Abstract

Objectives

To gain a comprehensive understanding of the vocal repercussions experienced by mild COVID‐19 infection, as well as the duration and underlying factors contributing to these effects.

Methods

Ten participants with mild COVID‐19 infection were included as the experimental group to evaluate the voice metrics changes at 15 days (D15), 30 days (D30), and 60 days (D60) after mild COVID‐19 infection, and 10 healthy people were taken as the normal control group. Self‐assessment and objective measures were taken at each time point, including questionnaires for voice handicap index (VHI) and reflux symptom index (RSI), as well as acoustic and aerodynamic indicators such as fundamental frequency (F0), Jitter, Shimmer, noise‐to‐harmonics ratio, sound pressure level (SPL), phonation threshold pressure (PTP), phonation threshold flow, aerodynamic resistance(AR), aerodynamic efficiency, mean expiration airflow, maximum phonation time, and maximum loudest phonation time.

Results

Notable elevations in RSI and VHI were observed during the D15 assessment, but by D60, these levels had returned to normal. At D60, Shimmer, PTP, and PTF significantly increased compared to the control group, while HNR and AR significantly decreased (p < 0.05). F0, SPL, and PTP significantly increased compared to 15 days (p < 0.05), while other acoustic and aerodynamic indicators showed no significant changes in intra‐group and inter‐group comparisons.

Conclusions

Our study demonstrated that COVID‐19 infection has a detrimental effect on voice production. Although subjective discomfort may gradually subside within 2 months after onset, alterations in phonation mode can be delayed. It is essential for healthcare professionals to remain vigilant in preventing any organic changes that may arise due to COVID‐19‐related voice disorders, such as muscle tension dysphonia.

Key points

• Voice sequelae of mild COVID‐19 will be delayed.

• Subjective and objective assessment of voice function in mild COVID‐19.

INTRODUCTION

The infection caused by the acute coronavirus‐2 (SARS‐CoV‐2) has been shown to elicit a systemic inflammatory response, which increases the risk of multiorgan system damage. ¹ , ² , ³ Researchers have conducted extensive studies on the post‐infection changes observed in the pulmonary, cardiovascular, renal, and digestive systems, as well as the mental health of patients and healthcare workers, ⁴ , ⁵ which have significantly contributed to our understanding of the pathophysiology and clinical manifestations of COVID‐19. ⁶

It has been noted that COVID‐19 can harm the organs responsible for voice production, including the lungs, larynx, nose, and throat. ⁷ , ⁸ , ⁹ Ren found that symptoms such as pharyngalgia and expectoration exist in different areas of Wuhan, a high‐incidence area of COVID‐19. ¹⁰ Moreover, individuals suffering from severe dyspnea and pneumonia due to COVID‐19 may require endotracheal intubation, ¹¹ which can further compromise the vocal tract. Consequently, vocal complications are common, and they may persist even after COVID‐19 recovery. ¹² , ¹³ , ¹⁴

Despite research on the long‐term vocal consequences of severe COVID‐19 cases, limited attention has been given to the potential voice sequelae in outpatients with mild infections. The significant number of mild COVID‐19 patients who experience laryngeal discomfort or persistent voice issues in the ENT outpatient has yet to be thoroughly examined. While the likelihood of a resurgence of COVID‐19 as a worldwide pandemic is currently low, it is still prevalent in specific regions and spreading. With the prospect of future infections, it is imperative that research on the voice sequelae following mild COVID‐19 infections become a priority. A limited number of studies rely solely on subjective scales and fail to incorporate objective evaluations in their exploration of the effects of mild COVID‐19 infections on voice function. ¹² , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ As a result, additional research is essential to understand the impact of mild illness on voice function, encompassing the length and origin of vocal sequelae.

Therefore, this study tracked the voice recovery phase after COVID‐19 infection and quantified the voice acoustic and aerodynamic changes, aiming to identify any potential long‐term effects on voice and associated complications.

METHODS

A follow‐up study was carried out in Guangdong Provincial People's Hospital from December 2022 to March 2023. Approval from the Internal Review Board (IRB) of Guangdong Provincial People's Hospital (No. KY‐Q‐2022‐032‐02) was obtained.

Subjects

The experimental group of this study consisted of 10 subjects, five females and five males, with an average age of 29.9 years. Select 10 healthy examinees (with an average age of 29.4 years) from our hospital as the normal control group.

Individuals aged between 20 and 40 who had tested positive for SARS‐CoV‐2 through RT‐PCR testing for COVID‐19 were recruited. The enrollment process was limited to patients with mild illness who have any of the various signs and symptoms of COVID‐19 (e.g., fever, cough, sore throat, malaise, headache, and muscle pain) but do not have shortness of breath, dyspnea, or abnormal chest imaging. Those who showed evidence of lower respiratory disease during clinical assessment or imaging and hospitalized COVID‐19 patients were excluded from the study. Additionally, individuals with a history of throat disease, neck surgery, or laryngopharyngeal reflux and those struggling with tobacco addiction were also excluded from the study. Excluded individuals with high vocal demands (such as teachers, singers, etc). The control group consisted of healthy individuals who underwent physical examinations in our department, denied any history of pharyngeal diseases or related laryngeal surgeries, and had no acute inflammation during the testing period. Functional and organic diseases of the vocal folds were excluded through a fibro laryngoscope. Before the study, all participants provided written informed consent. Experimental group subjects were all diagnosed with COVID‐19 for the first time and administered voluntary symptomatic treatment, which included antipyretics, painkillers, cough suppressants, and expectorants. Around the 9th day after symptom onset (9.0 ± 1.6 day), all patients tested negative for SARS‐CoV‐2.

Data collection

This experiment evaluated the self‐assessment, acoustic and aerodynamic indicators of the control group subject, as well as the voice indicators of the experimental group subjects at three‐time points: 15 days (D15), 30 days (D30), and 60 days (D60) after symptom onset. Each time point required the completion of self‐assessment, acoustic, and aerodynamic indicators to get a comprehensive view of each participant's vocal health.

Participants were instructed to keep track of their primary symptoms during each evaluation. They can document one or more symptoms they were experiencing in their current state. Additionally, the Voice Handicap Index (VHI) and Reflux Severity Index (RSI) were employed as subjective measures in our assessment. Participants were instructed to complete the forms based on their current condition.

The voice signal was captured utilizing a Shure microphone within a soundproof chamber. Participants were instructed to produce sustained vowel sounds, specifically/a/, three times at a comfortable pitch and volume. Acoustic parameters were analyzed utilizing Praat (Version 6.1.38), which included Jitter, shimmer, harmonic‐to‐noise ratio (HNR), sound pressure level (SPL), and fundamental frequency (F0). The statistical analysis was conducted based on the mean values of the acoustic parameters.

The Phonatory Aerodynamic System Model 6600 (KayPENTAX, Montvale, New Jersey) was used to analyze the aerodynamic parameters, including phonation threshold pressure (PTP), phonation threshold flow (PTF), airflow resistance (AR), airflow efficiency (AE), mean expiration airflow (MEA), maximum phonation time (MPT), and maximum loudest phonation time (MLPT).

The participants wore a face mask with an oral tube inserted 2 cm into their mouth to obtain the measurements of PTP and PTF. They were instructed to articulate the/pi/sound as softly as they could at a conversational pitch for three trials, and each trial was composed of five/pi/syllables. The mean value of the three central/pi/syllables underwent analysis after disregarding the initial and concluding syllables due to their instability and divergence. During the evaluation of AR and AE, the subjects were instructed to three trials for a total of 15/pi/syllables at a conversational pitch and volume when an oral tube was inserted 2 cm into the mouth. AR and AE were calculated based on the mean value of the three middle/pi/syllables across the three trials. The participants were instructed to utter the vowel/a/thrice while sporting a properly seated mask covering the nose and mouth to obtain MEA data. For MPT measurement, patients were asked to sustain an/a/in their normal voice for as long as possible. The MPT was determined by measuring the maintained/a/in three productions based on the voice signal. The longest sustained phonation was used for further processing. The patient was required to sustain an/a/in their loudest voice for as long as possible to acquire MLPT.

Statistical analysis

SPSS 23.0 (IBM) was used for statistical analysis. Data were tested for normality using the Shapiro–Wilk test. When the data passed the Shapiro–Wilk test, a one‐way analysis of variance (ANOVA) with repeated measures was performed to compare the data at the three‐time points. When the ANOVA showed significant differences (p < 0.05), a Tukey post hoc test was computed on the main significant effects or the interaction to compare the means. The comparison between the experimental and control groups was conducted using t‐test analysis method. If the data does not pass the Shapiro–Wilk test, perform a non‐parametric test.

RESULTS

On Day 15 of monitoring, seven patients reported coughing and throat clearing as their primary symptoms, while two patients reported hoarseness, and one expressed concern about expectoration. However, at Day 30, only two patients reported throat clearing as their primary symptom, and by Day 60, no notable signs were recorded, as indicated in Figure 1.

Figure 1.

The proportion of primary symptoms in patients with mild COVID‐19 infection at 15 days (D15), 30 days (D30), and 60 days (D60) after symptom onset.

Regarding the acoustic indicators, the results within the group showed there were significant differences in F0 and SPL among three‐time points (p = 0.017, p < 0.001, respectively), while no significant changes were found in Jitter, Shimmer, and HNR. An increased F0 was acquired at D60 compared to D15 in F0 (p = 0.001, Figure 2). SPL significantly increased at D30 and D60 compared with D15 in Figure 2 (p = 0.012, p < 0.001, respectively). The inter‐group results (Table 1) showed that the HNR of D60 was significantly reduced compared with the control group, and Shimmer was significantly increased (p = 0.008, p = 0.004, respectively).

Figure 2.

Comparison of F0, SPL, PTP, VHI, and RSI in patients with COVID‐19 mild infection at 15 days (D15), 30 days (D30), and 60 days (D60) after symptom onset. (*p < 0.05). F0, fundamental frequency; PTP, phonation threshold pressure; RSI, reflux symptom index; SPL, sound pressure level; VHI, voice handicap index.

Table 1.

Comparisons of acoustic indicators.

Group	F0 (Hz)	Jitter (%)	Shimmer (%)	HNR	SPL (dB)
Control	248.650 ± 66.879	0.226 ± 0.100	2.285 ± 0.979	25.098 ± 1.861	83.837 ± 7.318
D15	206.416 ± 63.628	0.256 ± 0.117	3.893 ± 1.583	17.374 ± 5.152	81.920 ± 7.999
D30	219.926 ± 65.182	0.196 ± 0.078	3.184 ± 1.170	18.996 ± 2.939	88.350 ± 9.067*
D60	227.170 ± 66.163*	0.235 ± 0.197	3.835 ± 1.997▲	19.473 ± 3.463▲	91.836 ± 9.067*

Note: Compared with the control group, ^▲ p < 0.05; compared to D15, *p < 0.05.

Abbreviations: HNR, harmonic to noise ratio; SPL, sound pressure level.

The intra‐group aerodynamics results show a statistically significant difference was observed in PTP between the three‐time intervals (p = 0.007). A rise in PTP was noted at D60 compared to D15, as illustrated in Figure 2 (p = 0.002). However, no significant differences were found in PTF, AR, AE, MEA, MPT, or MLPT. The inter‐group results showed (Table 2) that compared with the control group, the PTF of D15 was significantly increased (p = 0.009), the MPT of D30 was increased considerably (p = 0.002), while the PTP and PTF of D60 were significantly increased, and AR was reduced considerably (p = 0.005, p = 0.019, p = 0.001, respectively).

Table 2.

Comparisons of aerodynamic indicators.

Group	PTP (cmH2O)	PTF (l·s−1)	AR (cmH2O/[l·s−1])	AE (ppm)	MEA (l·s−1)	MPT (s)	MLPT (s)
Control	2.893 ± 1.024	0.031 ± 0.047	206.297 ± 160.870	535.324 ± 252.938	0.100 ± 0.0121	15.246 ± 4.400	18.425 ± 6.736
D15	3.111 ± 1.356	0.042 ± 0.042▲	128.953 ± 184.024	289.775 ± 176.615	0.066 ± 0.047	19.930 ± 5.964▲	19.697 ± 8.328
D30	3.466 ± 1.319	0.047 ± 0.101	130.446 ± 167.656	339.918 ± 331.124	0.070 ± 0.038	21.848 ± 9.094	21.216 ± 8.574
D60	5.124 ± 2.388▲	0.164 ± 0.174▲	81.800 ± 64.133▲	510.504 ± 381.206	0.148 ± 0.123	20.693 ± 6.866	20.235 ± 6.624

Note: Compared with the control group, ^▲ p < 0.05; compared to D15.

Abbreviations: AE, aerodynamic efficiency; AR, aerodynamic resistance; MEA, mean expiration airflow; MLPT, maximum loudest phonation time; MPT, maximum phonation time; PTF, phonation threshold airflow; PTP, phonation threshold pressure.

Within‐group results show significant differences were acquired in RSI and VHI among three‐time points (p = 0.001, p = 0.007). Notably, at D60, RSI and VHI levels decreased to nearly normal, as illustrated in Figure 2 (p < 0.001 and p = 0.005, respectively). The inter‐group results showed (Table 3) that compared with the control group, the RSI and VHI of D15 and D30 were significantly increased (p < 0.001 and p = 0.001 for D15, p = 0.034 and p = 0.007 for D30, respectively).

Table 3.

Comparison of self‐assessment scale scores.

Group	VHI	RSI
Control	0.500 ± 0.417	0.923 ± 1.244
D15	7.900 ± 72.656▲	11.700 ± 87.600▲
D30	3.300 ± 5.567▲	4.000 ± 12.700▲
D60	1.200 ± 1.733*	1.700 ± 7.790*

Note: Compared with the control group, ^▲ p < 0.05; compared to D15, *p < 0.05.

Abbreviations: RSI, reflux symptom index; VHI, voice handicap index.

DISCUSSION

Our study demonstrated that COVID‐19 infection negatively impacted voice production, which could be long‐lasting.

This experiment used two self‐assessment scales. Previous studies have reported the possibility of laryngopharyngeal reflux during the COVID‐19 recovery period. ¹⁹ The reflux symptom index (RSI) questionnaire is an effective tool for assessing reflux and can provide valuable insights into whether COVID‐19 infection accelerates laryngopharyngeal reflux. VHI is a widely accepted scale for voice assessment, which can determine voice outcomes in patients who have recovered from COVID‐19. Therefore, we included RSI and VHI in this research.

The subjective symptoms of the patient are most obvious at D15. Coughing and throat clearing were the most frequently reported symptoms identified by 70% of patients; the average RSI and VHI values are significantly higher than the normal values at this time. Consistent with prior research, ²⁰ this study found that two patients exhibited RSI scores that surpassed the threshold of 13 at D15, suggesting a potential risk for LPR. Previous research has found a similar outcome in that the diagnoses of LPR were common in non‐intubated post‐COVID patients. ²¹ , ²² However, the underlying reason behind this issue is still uncertain. It is possible that COVID‐19 infections may lead to an increase in RSI as a result of inflammation. ACE‐2 receptors, which are the entry point for the virus, are present in the upper aerodigestive tract. ²¹ As a result, the upper aerodigestive tract may also be affected by SARS‐CoV. COVID patients with impaired laryngeal mucosa may display a heightened sensitivity to alterations in pH levels. The variation in mucosa tissue due to inflammation may also cause the increased VHI. Additionally, the rise in the VHI score is primarily attributed to the decline in the functional and physical aspects rather than the emotional part. The findings suggested that voice alterations may adversely impact sound emission, but the resulting temporary voice impairment does not induce emotional instability. On Day 15, the patient's subjective symptoms increased, and the deterioration of self‐assessment indicators and PTF suggested that infection had a certain impact on vocal function. Two patients experiencing dysphonia within the initial 2 weeks of observation exhibited a gradual reduction in jitter and shimmer over time. However, no statistically significant differences were detected in jitter and shimmer, as well as HNR. These results were consistent with the findings in Golac et al.'s study. ²⁰ The vocal fold has been found to be associated with ACE2 expressio, ¹² but levels of ACE2 may vary between individuals. Despite the impact of the inflammatory response on the vocal tract mucosa, we hypothesized that there were few adverse effects on the mucosal wave of the vocal folds.

Although the SPL of patients increased significantly at D30, and the values of related self‐assessment indicators did not ultimately return to normal, the subjective symptoms of patients decreased significantly, and other acoustic indicators were not significantly abnormal, which indicated that COVID‐19 infection had entered the recovery period. This result suggests that the patient's vocal style may change with the relief of throat symptoms. Considering the aerodynamic changes of D60 in this study, it is possible that this may indicate the patient's potential for vocal hyperfunction.

After 2 months of infection, both RSI and VHI returned to normal levels, but surprisingly, PTP, F0, and SPL increased significantly compared to before. The experimental group also showed significant aerodynamic changes in PTP and PTF, and AR decreased significantly compared to the control group at D60. The cause behind it warrants investigation. Combined with our analysis of subjective assessment and acoustic metrics, it appears that patients have trouble controlling their vocalizations during recovery. As their symptoms relieve, patients tend to misuse their laryngeal muscles, causing increased tension in their vocalization and higher F0 and SPL levels. Patients also exhibit a reduced ability to control their voice when speaking softly, leading to a higher PTP. These findings indicate a potential risk of muscle tension dysphonia (MTD) for patients. This outcome is consistent with the result of Allisan‐Arrighi et al., who reported that 38% of non‐intubated patients were diagnosed with MTD. ²¹ This reminded us that even for mild COVID‐19 infection, more attention should be taken to long‐term voice sequelae. Therefore, healthcare practitioners should stay alert and prevent any organic changes that may arise due to MTD.

The cause of voice sequelae following COVID‐19 remains elusive. It is unclear whether the observed symptoms are a result of specific voice issues caused by COVID‐19 or common side effects of cough and laryngitis subsequent to acute laryngitis. Despite the prevalence of acute laryngitis, it is frequently managed by primary physicians, and related voice measurements are scant. Jaworek et al. ²³ conducted a study comprising seven acute laryngitis patients and found that these individuals frequently presented accompanying upper respiratory tract infections with symptoms such as hoarseness, sore throat, dysphagia, dyspnea, cough, congestion, and postnasal drip. The voice may sound breathy and/or raspy, which typically lasts for 3–8 days in most cases. These symptoms usually subside within 3 weeks after the onset of acute laryngitis. However, our observation of the potential risk of MTD occurs in D60. No research has yet monitored acoustic and aerodynamic voice changes over 2 months following an acute respiratory infection. Nevertheless, based on our clinical experience, fewer respiratory infection patients have reported voice dysphonia or have been diagnosed with MTD, which is significantly lower than the previously reported 38% of COVID‐19 patients with MTD. Therefore, we believe that the voice function change in this study may be a specific voice problem caused by COVID‐19 infection, which needs to be paid attention to and distinguished by healthcare professionals, but this view needs more clinical and basic research to confirm.

This study has limitations due to a lack of available voice metric data prior to infection. Due to this experiment's small sample size and limited follow‐up time, individual differences should be considered. Our measurements were collected during the peak of the outbreak in China when many individuals were in quarantine, and no more patients could be recruited for the experiment. Although the sample of this study is small, all the patients were young, healthy adults who were infected during the same wave of the pandemic, making this study a homogeneous sample that yields pure results.

CONCLUSION

The study involved healthy young adults who presented with mild COVID‐19 symptoms. This study shows that although the subjective discomfort of patients with mild COVID‐19 infection will gradually subside within 2 months after the onset, the voice function changes during the recovery period of COVID‐19 infection may cause changes in pronunciation patterns, which may lead to functional pronunciation disorders.

AUTHOR CONTRIBUTIONS

Zhi‐Xue Xiao conducted the experiment and wrote the manuscript. Qing‐Yi Ren collected the data and helped experiment. Wei‐Qing Liang done data collection, Na‐Na Li done data analysis. Nan Huang done statistical analysis. Zhi‐Xian Zhu revised the manuscript. Ping‐Jiang Ge recruited patients and revised the manuscript. Si‐Yi Zhang designed the experiment and recruited patients. Jing Kang designed and conducted the experiment, revised the manuscript.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

ETHICS STATEMENT

This study was approved by the Internal Review Board (IRB) of Guangdong Provincial People's Hospital (No. KY‐Q‐2022‐032‐02).

Xiao Z‐X, Ren Q‐Y, Liang W‐Q, et al. Follow‐up of patients with mild COVID‐19 using subjective, acoustic, and aerodynamic measurements. World J Otorhinolaryngol Head Neck Surg. 2025;11:400‐405. 10.1002/wjo2.234

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request due to privacy or ethical restrictions.