Related Experiment Video
Updated: Sep 18, 2025
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
An unsupervised cluster analysis of multimorbidity patterns in older adults in Shenzhen, China
Xiaolong Guo1, Peiyi Liu2, Jing Guo3
1School of Biomedical Engineering, Shenzhen University Medical School, Shenzhen University, Shenzhen, China.
Background:
Population aging challenges health care systems due to the high prevalence and impact of multimorbidity in older adults. Studies on multimorbidity in Shenzhen have primarily focused on the quantity of multimorbidity, lacking in-depth exploration of multimorbidity patterns.
Methods:
Based on baseline data from the Shenzhen aging-related disease cohort, this study analyzed information from 8,911 people aged 60 and above after excluding missing and abnormal values from interview results. Using self-organizing map combined with weighted k-means, the distribution of diseases in the population was visualized, dividing the overall population into four clusters. The study also analyzed comorbidity and association rules for each cluster.
Result:
This study found a high prevalence of cardiometabolic comorbidities among the older adult in Shenzhen, reaching 15.83%, and detailed the distribution of specific comorbidity combinations. Hypertension had a high prevalence and was the most common factor in comorbidities among Shenzhen's older adult. Additionally, hyperuricemia was included as a disease to explore its multimorbidity patterns with other chronic conditions.
Conclusion:
The study found that multimorbidity is prevalent among the older adult in Shenzhen and explored their patterns, suggesting that Shenzhen should enhance screening and integrated management of high-risk groups and implement public health interventions to alleviate the multimorbidity burden.
