A laryngeal disorders prediction model based on cluster analysis and regression analysis

Abstract

This study provided the baseline for establishing policies for community health promotion programs to propose the clusters of multiple health risk factors and identify the risks of laryngeal disorders according to the clusters by using the national level survey representing the South Korean population. This study targeted 5941 people who completed the 5th Korean National Health and Nutrition Examination Survey. The independent variables were age, sex, smoking, high-risk drinking, education level, occupation, household income, and self-reported voice problems. The identify cluster relationship with laryngeal disorders by conducting 2-way cluster analysis and multinomial logit analysis. The prevalence of laryngeal disorder was 6.7%. The results of analysis, 3 clusters were automatically extracted. Multinomial logistic regression analysis showed that sociodemographic factors, health risk behaviors, and health status clusters were significantly related to the risk of laryngeal disorders. The males who were smoking, high-risk drinking, college graduate and above, high income, and non-manual workers had a higher risk of laryngeal disorders than females who were non-smokers, non-drinkers, 60 years old and older, economically inactive, and high school graduate. The results of this study suggested that it may be effective to classify population according to sociodemographic and health behaviors and develop health education materials and health promotion program accordingly in order to prevent laryngeal disorders.

Predicting High-Risk Groups for COVID-19 Anxiety Using AdaBoost and Nomogram: Findings from Nationwide Survey in South Korea

“…It is known that the prevalence of anxiety disorders is higher in women than in men [29], in the young and the prime-aged than the elderly [30], in the people with a lower education level than those with a higher education level [22,31,32], and in those with depressive disorders [33]. However, this study found that older adults felt more anxious than those in their prime age.…”

“…However, some diseases may violate the assumption of this regression model. Moreover, since risk factors for diseases are complex, risk factors are likely to cluster with each other, and the effects of risk factors on the health level are synergistic rather than additive [22], it is challenging to identify multiple risk factors for anxiety using the regression model alone.…”

Exploring the Predictors of Rapid Eye Movement Sleep Behavior Disorder for Parkinson’s Disease Patients Using Classifier Ensemble

“…Until now, various fields including health science have used traditional statistical (classification) techniques (e.g., multiple logistic regression analysis and multiple discriminant analysis) and various algorithms (e.g., artificial neural network and decision tree) as machine learning techniques for predicting target variables [16]. However, these algorithms may have the following limitations depending on the characteristics of the dataset: (1) although regression analysis is effective for identifying individual risk factors, the selection of risk factors depends on researchers' rule of thumb [17]; (2) artificial neural network and decision tree are vulnerable to overfitting, noise, and outliers [18]; and (3) the accuracy of decision trees is highly likely to vary depending on the type and quantity of input variables.…”

Is Deep Learning Better than Machine Learning to Predict Benign Laryngeal Disorders?

“…Consequently, it is important to fully understand the etiology of a benign laryngeal mucosal disorder and identify multiple risk factors of the disease in order to perform accurate diagnosis and treatment. Nevertheless, most studies that have evaluated the risk factors of laryngeal disorders have just tried to find individual risk factors using regression analysis [25,26,27,28,29], and only a few studies have explored the multiple risk factors of benign laryngeal mucosal disorders using machine learning [30].…”