Epidemiological exploration of the impact of bluetooth headset usage on thyroid nodules using Shapley additive explanations method

Authors: Zhou N, Qin W, Zhang JJ, Wang Y, Wen JS, Lim YM

Year: 2024 Jun 21

Category: Epidemiology

Journal: Sci Rep

DOI: 10.1038/s41598-024-63653-0

URL: https://www.nature.com/articles/s41598-024-63653-0

Abstract

Overview

With thyroid nodules becoming increasingly prevalent worldwide, this study investigates the potential correlation between Bluetooth headset use and thyroid nodule incidence. Special focus is given to the cumulative effects of non-ionizing radiation (NIR) emitted by Bluetooth devices.

Methods

  • Analyzed 600 valid questionnaires from the WenJuanXing platform.
  • Applied Propensity Score Matching (PSM) and an XGBOOST machine learning model, enhanced with SHAP (Shapley Additive Explanations) analysis to assess risk factors.
  • PSM balanced baseline characteristics to reduce bias, resulting in a matched dataset of 96 cases for in-depth modeling analysis.
  • The XGBOOST model achieved an Area Under the Curve (AUC) of 0.95, indicating high prediction accuracy for thyroid nodule risk factors.

Findings

  • SHAP analysis identified age and daily Bluetooth headset usage duration as the two most significant factors impacting thyroid nodule risk.
  • Notably, longer daily usage durations of Bluetooth headsets showed a strong association with increased risk of developing thyroid nodules.

Conclusion

This study highlights a significant relationship between prolonged Bluetooth headset use and increased thyroid nodule risk. The findings underscore the need to consider potential health impacts when using modern wireless devices and suggest public health guidelines should address limiting daily Bluetooth headset duration.

The research also calls for future investigations into the biological mechanisms underlying this association and urges the development of comprehensive health guidance and preventive measures regarding Bluetooth headset exposure.

Bluetooth headset thyroid nodules non-ionizing radiation SHAP analysis XGBOOST risk factors public health
← Back to Stats