Diagnostic Value of Acoustic Analysis and Inflammation Markers in Pediatric Vocal Fold Nodules

Introduction

Vocal fold nodules (VFN) are small, firm tissue lesions that develop on the vocal folds, typically as a result of excessive or improper voice use. These lesions are typically symmetrical and bilateral. The primary cause of vocal nodules is repetitive trauma to the vocal folds [1]. Such trauma is particularly common among teachers, singers, and individuals who use their voices extensively. Microinjuries resulting from this trauma disrupt the mucosal barrier, initiating inflammatory processes in the local tissues. This leads to changes in the vibratory behavior of the vocal folds, resulting in aperiodic vibration and, consequently, a reduction in voice quality. The sex disparity ratio of VFN occurrence is approximately 2: 1, more commonly seen in males [2]. The incidence of VFNs among school-aged children has been reported as 16.9%, in which male patients have a higher prevalence (21.6%) compared to female patients (11.7%) [3].

Objective and subjective methods are utilized to evaluate the degree of dysphonia and the acoustic quality of the voice. Objective voice analysis is a noninvasive technique used to investigate and understand the physical characteristics of vocal output. The fundamental frequency (F0), which reflects the number of vocal fold vibrations per second, is often reduced in individuals with vocal nodules due to increased stiffness and mass of the vocal folds [4]. Elevated jitter values indicating irregular variations in pitch are frequently observed in patients with VFN as a result of asymmetric vocal fold vibrations [5]. Similarly, shimmer, which measures variations in vocal intensity, tends to be higher in VFN cases due to the irregular vibratory patterns caused by the nodules [6]. Another important parameter, the noise-to-harmonics ratio (NHR), quantifies breathiness in the voice, a characteristic often seen in VFN due to incomplete closure of the vocal folds [7]. Higher NHR values are typically associated with the breathy quality of the voice, which results from inefficient vocal fold closure. A study conducted by Aoki et al investigated the acoustic properties of the voice in pediatric patients with VFN [8]. The findings demonstrated that vocal nodules impair vocal fold vibration, leading to reduced control over pitch and loudness and, consequently, an increase in noise components in the voice signal.

Blood-based inflammatory markers are critical for identifying systemic inflammation. The systemic immune inflammation index (SII), mean platelet volume (MPV), platelet-to-lymphocyte ratio (PLR), and neutrophil-to-lymphocyte ratio (NLR) are widely-used inflammatory markers [9]. These markers are important as prognostic values in several conditions, including solid malignancies, inflammatory disorders, and Bell’s palsy [10,11]. There are studies in the literature showing the relationship between vocal nodules and inflammation. However, a literature search found no study showing the relationship with blood inflammatory parameters [12,13].

Acoustic parameters, while reflective of the current vocal status, are not disease-specific, as numerous lesions and dysfunctions can induce changes in their values. Considering that the primary mechanism underlying the development of VFNs is mechanical trauma followed by inflammation, early control of the inflammatory process may contribute to regression of these lesions. Based on this hypothesis, the present study was designed to explore the potential relationship between acoustic voice parameters and systemic inflammatory markers, with the goal of contributing to early diagnosis and informing therapeutic strategies. Controlling inflammation at an early stage may facilitate the resolution of VFNs. Therefore, this study aimed to investigate the effect of systemic inflammation as a potential etiological factor in VFN, using acoustic parameters as a complementary tool. Additionally, the study sought to establish a possible diagnostic correlation between VFN and inflammatory parameters, including PLR, SII, and NLR.

Material and Methods

STATISTICAL ANALYSIS:

The Shapiro-Wilk test was used to assess the normality of data distribution. Comparisons between groups with quantitative variables that follow a normal distribution were carried out using the independent samples t test. The Mann-Whitney U test was used to analyze variables that did not follow a normal distribution. Group comparisons for the normally distributed quantitative variables NLR, PLR, and MPT were performed using the independent samples t test. For the variables that were not normally distributed – age, F0 Mean, F0 Min, F0 Max, SD, Jitter, Shimmer, Mean NHR, and pVHI-10 – the Mann-Whitney U test was used. Receiver operating characteristic (ROC) analysis was conducted to determine which parameters that best distinguish the disease, and the area under the curve (AUC) was calculated. AUC values were recorded, and specific cut-off points were identified. Statistical parameters are presented as Mean±SD, Median (Q1–Q3), and n (%). A p value less than 0.05 was considered statistically significant. Data analysis was conducted using IBM SPSS Statistics version 22.

Results

The VFN group and control group were matched for age and sex. Among the VFN group, 19 participants were male and 11 were female, while the control group included 20 males and 10 females. The mean age of the VFN group was 10.00 years (range: 8.00–13.00), and the mean age of the control group was 11.50 years (range: 8.00–14.00).

Acoustic analysis revealed the following mean values for the examined parameters in the VFN group: F0=262.04 Hz, jitter (%)=0.60, shimmer (%)=4.30, NHR=0.02, and SD=3.85. In the control group, the values were F0=251.45 Hz, jitter (%)=0.31, shimmer (%)=1.49, NHR=0.00, and SD=1.73. Comparisons between the groups showed that the VFN group had significantly higher jitter, shimmer, NHR, and SD measurements. MPT was 8.08±2.24 seconds in the VFN group and 9.94±2.82 seconds in the control group, showing a statistically significant decrease in the VFN group (p=0.006). The pVHI was calculated as 11.00 (range: 9.00–15.00) in the VFN group and 1.00 (range: 0.00–2.00) in the control group, with significantly lower scores observed in the control group (p<0.001) (Table 1).

According to the ROC analysis results, the F0 max variable demonstrates a moderate level of discrimination between the VFN group and the control group. The cut-off point is the best value that distinguishes between diseased and healthy individuals. The optimal cut-off point for F0 max, offering the best discrimination between the VFN group and control group, was 265.04, with both sensitivity and specificity values of 0.667. This cut-off point was found to be statistically significant (p=0.016). The jitter variable exhibited a high level of discrimination, with a sensitivity of 0.767 and specificity of 0.833. This discrimination was statistically significant (p<0.001). Similarly, the shimmer variable also demonstrated a high discriminative ability, with a sensitivity of 0.967 and specificity of 0.733. The cut-off point for shimmer was statistically significant (p<0.001). Likewise, the mean NHR and VHI-10 variables showed strong discriminatory power, with statistically significant cut-off points (p<0.001 for both). The MPT variable demonstrated a moderate level of sensitivity, with a sensitivity of 0.733 and specificity of 0.633. This variable also had a statistically significant cut-off point (p=0.005) (Figure 1). The sensitivity and specificity rates of the cut-off values are shown in Table 2.

NLR, PLR, MPV, and SII were lower in the VFN group, with the decrease in NLR and PLR values being statistically significant (p=0.010 and p=0.019, respectively) (Table 3). In the ROC analysis, the area under the curve (AUC) for NLR values was higher than that for PLR, MPV, and SII values (Figure 2). The NLR variable was at the cut-off point 1.18, sensitivity value 0.667, and the specificity value 0.667. NLR has statistically significant and important discrimination ability (p=0.012). It exhibits a moderate discrimination performance (AUC=0.689). PLR, MPV, and SSI variables did not provide a significant cut-off point in terms of discrimination (Table 4). This analysis shows that the NLR value is a good potential marker in patients with VFN. Therefore, NLR has greater specificity and positive predictive value than other measurements.

Discussion

Although the etiology of VFN disorders remains unknown, they have been associated with recurrent mucosal injuries caused by laryngeal hyperfunction and vocal abuse. This study explored the correlation between inflammatory conditions and vocal nodules by performing voice analyses on affected patients. Acoustic analyses revealed that F0 max, SD, jitter, shimmer, NHR, MPT, and pVHI values were significantly altered in patients with VFN. Analysis of the cut-off values for these parameters demonstrated that they had diagnostic significance. Additionally, several inflammatory markers, including NLR and PLR, were found to be significantly lower. Specifically, the cut-off values of the NLR (1.18/0.689) were significant for progression of the disease.

Acoustic voice analysis, which evaluates parameters such as frequency, intensity, and periodicity of the voice, is essential for the diagnosis and follow-up of voice disorders. Wilson reported that children with vocal fold nodules often speak 3 times more than their healthy counterparts [16]. According to the voice analysis results in the present study, the F0 value was inversely related to age in both groups, although this difference was not statistically significant. Wertzner et al reported no statistically significant variances in F0, jitter, and shimmer values, while Niedzielska et al identified statistically significant variances in jitter, shimmer, and NHR values but not in F0 [17,18]. Previous studies have reported both lower and higher F0 values following therapy, but these studies included data from patients who did not fully recover [19,20]. In the present study, F0 values were higher in the VFN group compared to healthy controls, although the difference was not statistically significant. This discrepancy was attributed to variations in voice pitch related to age and sex in both groups. Aoki et al investigated the sensitivity and specificity of the voice parameters PPQ, APQ, and NHR in predicting VFN in children, showing that NHR has high specificity for this purpose. Similarly, in the present study, NHR also showed high specificity, and the findings were consistent with those of Aoki et al. In a study by Akif et al, the F0 value was not significant in the voice analysis of patients with VFN [3]. In contrast, the present findings demonstrated that while the mean F0 value was not significant, several parameters, including F0 max, SD, jitter, shimmer, and NHR, were significant. Furthermore, the diagnostic value of the cut-off points for these parameters was also confirmed.

MPT is a valuable measure for assessing voice health and monitoring treatment progress. Prolongation of MPT has been observed in adductor spasmodic dysphonia with severe glottic closure, whereas shortening has been attributed to glottic insufficiency, pulmonary insufficiency, or submaximal effort. The present study found that MPT was significantly shorter due to the closure pattern disorder observed in every patient with nodules. Zur et al stated that the pVHI is useful for assessing the effect of voice issues in pediatric populations, as well as evaluating their condition following surgery, treatment, or voice therapy [21]. There are many studies in the literature showing the reliability of pVHI in various languages. These studies support that it is a reliable tool across different age groups and based on parental assessments [22,23]. Similarly, the present study found the pVHI scores were significantly higher compared to the control group, confirming its utility as a subjective assessment method.

NLR is a measure that reflects the balance between 2 distinct immune pathways. Neutrophils serve as the initial defense in the immune system, displaying phagocytic and apoptotic actions by releasing various inflammatory mediators, particularly cytokines [24]. Lymphocytes act as specific inflammatory mediators with regulatory or protective functions; low lymphocyte counts are indicative of suboptimal overall health and physiological stress [25,26]. Studies have shown that people with voice nodules have higher levels of fear, are more likely to worry, and have more psychological and vocal distress [27,28]. Vocal nodule formation occurs as a result of increased vocal stress, and some inflammatory changes occur as a result of increased systemic stress. In addition, increased levels of NLR are noted in cases of widespread inflammation, particularly in various gynecological tumors, gastrointestinal malignancies, and cardiovascular diseases. It is also considered a potential marker for inflammatory conditions resulting from infections or autoimmune diseases [26,29]. In recent years, specific cut-off values for NLR have been identified as predictive in different types of cancer. The present study determined a cut-off value for NLR of 1.18, which suggests that NLR could be a distinguishing marker for identifying patients with VFN.

The PLR serves as a non-specific inflammatory marker, regulating platelet function, endothelial passage, and the activity of neutrophils and macrophages. PLR values tend to increase rapidly in inflammatory diseases [30]. Additionally, high PLR levels have been suggested as valuable markers for predicting local recurrences in tumoral tissues. Elevated PLR values have been reported in peripheral vascular problems, coronary artery diseases, and certain gynecological and hepatobiliary cancers, although their association with prognosis is weak [30,31]. The present study found significantly lower PLR values in patients with VFNs. This finding may be related to the higher lymphocyte ratios in these patients, reflecting an imbalance in their immune systems. Changes in the immune system may also cause changes in NLR and PLR values. The present study hypothesized that evaluation of voice parameters together with blood inflammatory values taken in the early stages of the disease could be used in early diagnosis or clinical applications during the disease process, and may contribute to more comprehensive studies in the future.

SII, calculated by multiplying ANC with platelets, has gained attention as a new inflammation marker, hypothesized to represent the equilibrium between the inflammatory and immune responses, and has been investigated as a marker of inflammation and to assess the prognosis in various conditions, including malignancies, asthma, nasal polyps vasculitis, and Bell’s palsy [32–34]. In the present study, although SII was elevated in the VFN group, the results were not statistically significant. Increasing the sample size in future studies may yield significant findings.

Lymphocytes play a crucial role in targeting and destroying tumor cells through cellular and humoral immune responses. An increase in their count indicates immune system activation [35]. The present study found a statistically significant decrease in NLR and PLR ratios in the VFN group. However, it may be that the increase in lymphocyte count, due to reduced neutrophil and platelet values, does not represent a true increase. These findings suggest that the results reflect immune system irregularities in patients with vocal fold disorders.

The limitations of this study include the retrospective design and its single-center sampling, leading to a smaller sample size. Additionally, the lack of a comparative analysis of the hematologic data before and after treatment limits the depth of the findings. Future studies involving post-treatment hematological parameters would provide more comprehensive and useful data.

Conclusions

This study focused on irregularities in blood parameters associated with the immune system. Specifically, the findings suggest that blood inflammatory markers may serve as potential biomarkers for VFNs, in line with evidence from the literature, and demonstrates the potential clinical relevance of these parameters in voice analysis for evaluation of VFNs. These findings may provide a guide for future research.