A recent study published in Engineering utilized machine learning to identify distinct clusters of chronic obstructive pulmonary disease (COPD) patients in China, highlighting how comorbidity profiles impact health-related quality of life (HRQoL). Conducted by researchers from the Chinese Academy of Medical Sciences & Peking Union Medical College, Heidelberg University, Stanford University, and other leading institutions, the study leverages data from the Chinese Enjoying Breathing Program to provide insights into the heterogeneity of COPD and its implications for targeted public health interventions.

COPD is a progressive respiratory disorder and a major global health burden, ranking as the fourth leading cause of death worldwide in 2021. The prevalence of COPD among Chinese adults aged 20 years and older was 8.6% in 2015, according to the China Pulmonary Health Study. The condition is characterized by persistent airflow limitation and is increasingly recognized as a heterogeneous disease, often complicated by comorbidities such as cardiovascular disease, asthma, bronchiectasis, and diabetes. These comorbidities significantly affect HRQoL, disease burden, and survival.

The study employed a cross-sectional, multicenter cohort design, using data from the Enjoying Breathing Program (2020–2023). The researchers included 11,145 COPD patients, of whom 6616 (59.36%) had at least one comorbidity. The primary outcome was HRQoL, measured using the EQ-5D five-level verison (EQ-5D-5L) instrument. The researchers used multiple correspondence analysis to reduce dimensionality from 31 variables, including 27 comorbidities and four socio-demographic or health-related characteristics, to three uncorrelated components. Unsupervised machine learning algorithms, specifically K-means++ and hierarchical clustering, were then applied to identify distinct clusters. (The Enjoying Breathing Program was registered at ClinicalTrials.gov (ID: NCT04318912) in March 2020. This study was approved by the China–Japan Friendship Hospital, China (Approval number: 2019-41-k29). All participants provided written informed consent. This study followed the Declaration of Helsinki ethical principles.

The analysis identified four distinct COPD patient clusters: young male smokers, biomass-exposed females, respiratory comorbidity, and elderly multimorbid. The young male smokers cluster was the largest and youngest, predominantly male, with the highest proportion of current and former smokers and the lowest comorbidity rate. The biomass-exposed females cluster was primarily female, with low smoking prevalence but high biomass fuel exposure. The respiratory comorbidity cluster was characterized by the lowest lung function and a predominance of chronic bronchitis and pulmonary emphysema. The elderly multimorbid cluster was predominantly aged 70 years and older, with high rates of hypertension, ischemic heart disease, and diabetes.

The study found that HRQoL declined with increasing comorbidities. The young male smokers cluster reported the highest EQ-5D-5L utility score (0.74), while the respiratory comorbidity and elderly multimorbid clusters reported the lowest scores (0.66 and 0.65, respectively). The respiratory comorbidity cluster exhibited the poorest overall outcomes, with significantly increased risks in mobility, activity, and anxiety/depression, followed by the elderly multimorbid cluster, which demonstrated worse HRQoL, particularly in mobility and pain/discomfort.

The findings underscore the need for targeted public health interventions and integrated care strategies for COPD management. By identifying distinct patient clusters, this study provides a nuanced understanding of COPD in China and informs the development of more effective public health and clinical approaches. Future research should validate these clusters and assess their applicability in guiding COPD-related policy and management approaches.

The paper “Exploring COPD Patient Clusters and Associations with Health-Related Quality of Life Using A Machine Learning Approach: A Nationwide Cross-Sectional Study,” is authored by Chao Wang, Fengyun Yu, Zhong Cao, Ke Huang, Qiushi Chen, Pascal Geldsetzer, Jinghan Zhao, Zhoude Zheng, Till Bärnighausen, Ting Yang, Simiao Chen, Chen Wang. Full text of the open access paper: https://doi.org/10.1016/j.eng.2025.05.005. For more information about Engineering, visit the website at https://www.sciencedirect.com/journal/engineering.