Machine learning insights into the burden and risk factors of lower respiratory infections in East Asia (1990–2033)
- 1Department of Respiratory Medicine, The Third People's Hospital of Chengdu, Affiliated Hospital of Southwest Jiaotong University, China
- 2Institute of Clinical Medicine, Shanghai Jiaotong University School of Medicine, China
- 3School of Global Health, Chinese Center for Tropical Diseases Research, Shanghai Jiao Tong University School of Medicine, China
Abstract
Objectives: This study quantified trends in the burden of lower respiratory infections from 1990 to 2023 in five East Asian countries, projected disease trajectories from 2024 to 2033 using machine learning, and identified key risk factors and their associations with the sociodemographic index.
Methods: Based on the Global Burden of Disease Study (2023) estimates, we calculated the age-standardized incidence, prevalence, mortality, and disability-adjusted life year rates of lower respiratory infections for five East Asian countries. A multistage machine learning framework with Shapley additive explanations was used to screen exposure-related risk factors, and machine learning time-series models were used to project the future burden of lower respiratory infections.
Results: Lower respiratory infection incidence and prevalence declined universally; however, mortality trends were heterogeneous (rising in South Korea and falling in others). China demonstrated the largest reduction in incidence. Mongolia demonstrated the steepest decline in prevalence; however, it continued to have a high burden of disability-adjusted life years in 2023. Exposure to secondhand smoke was the most important risk factor, whereas zinc/iron deficiency prevailed in low sociodemographic index settings; the lower respiratory infections burden had a nonlinear inverse association with the sociodemographic index, with population aging as a key driver. Projections demonstrated a continued decline in the overall burden of lower respiratory infections, although mortality rates were projected to increase in South Korea.
Conclusion: The burden of lower respiratory infections has reduced substantially in East Asia over three decades; however, population aging has resulted in divergent mortality trends. High-sociodemographic index regions require age-responsive, integrated strategies targeting both exposure reduction and host vulnerability to sustain declines in mortality rates.
Related
- Etiology and co-infection patterns of acute respiratory viral infections in Bhopal, India: insights from short-term molecular study
- Application of machine learning to identify risk factors for bloodstream infections caused by different antimicrobial resistance phenotypes in chemotherapy patients.
- Factors associated with discharge on higher respiratory support for children with tracheostomy hospitalized with bacterial tracheostomy‐associated infections: A prospective, multicenter cohort study
- Global burden of respiratory infections and tuberculosis and risk factors, 1990-2021: a systematic analysis of the Global Burden of Disease Study 2021.
- Meteorological forecasting of respiratory diseases incidence with geospatial and machine learning methods
BibTeX
@article{su2026machine,
title = {{Machine learning insights into the burden and risk factors of lower respiratory infections in East Asia (1990–2033)}},
author = {Xian Su and Wu Xu and Wenwen Lv and Jinxin Zheng and Yuebin Wang and Tingting Wu and Li Sun and Huirong Hu and Yiping Liu and Yaolin Li},
journal = {The Journal of International Medical Research},
year = {2026},
month = sep,
doi = {10.1177/03000605261467663},
url = {https://doi.org/10.1177/03000605261467663},
note = {CC BY-NC 4.0},
}
RIS
TY - JOUR
TI - Machine learning insights into the burden and risk factors of lower respiratory infections in East Asia (1990–2033)
AU - Xian Su
AU - Wu Xu
AU - Wenwen Lv
AU - Jinxin Zheng
AU - Yuebin Wang
AU - Tingting Wu
AU - Li Sun
AU - Huirong Hu
AU - Yiping Liu
AU - Yaolin Li
PY - 2026
DA - 2026/09/29
JO - The Journal of International Medical Research
DO - 10.1177/03000605261467663
UR - https://doi.org/10.1177/03000605261467663
AB - Objectives: This study quantified trends in the burden of lower respiratory infections from 1990 to 2023 in five East Asian countries, projected disease trajectories from 2024 to 2033 using machine learning, and identified key risk factors and their associations with the sociodemographic index. Methods: Based on the Global Burden of Disease Study (2023) estimates, we calculated the age-standardized incidence, prevalence, mortality, and disability-adjusted life year rates of lower respiratory infections for five East Asian countries. A multistage machine learning framework with Shapley additive explanations was used to screen exposure-related risk factors, and machine learning time-series models were used to project the future burden of lower respiratory infections. Results: Lower respiratory infection incidence and prevalence declined universally; however, mortality trends were heterogeneous (rising in South Korea and falling in others). China demonstrated the largest reduction in incidence. Mongolia demonstrated the steepest decline in prevalence; however, it continued to have a high burden of disability-adjusted life years in 2023. Exposure to secondhand smoke was the most important risk factor, whereas zinc/iron deficiency prevailed in low sociodemographic index settings; the lower respiratory infections burden had a nonlinear inverse association with the sociodemographic index, with population aging as a key driver. Projections demonstrated a continued decline in the overall burden of lower respiratory infections, although mortality rates were projected to increase in South Korea. Conclusion: The burden of lower respiratory infections has reduced substantially in East Asia over three decades; however, population aging has resulted in divergent mortality trends. High-sociodemographic index regions require age-responsive, integrated strategies targeting both exposure reduction and host vulnerability to sustain declines in mortality rates.
SN - 0300-0605
SN - 1473-2300
ER -