Abstract
Background: Multimorbidity poses major healthcare challenges which contributes to a decline in quality of life and an increased mortality risk. There exists heterogeneity on the internal associations within multimorbidity. We aimed to explore multimorbidity patterns and construct networks, delving into the relationships among diseases.
Methods: The data from the health examination records of adults residing in Southern China in 2020 were utilized. Individuals aged 65 and above were included. Fifteen diseases were extracted. Hierarchical cluster analysis was performed. The multimorbidity matrix was calculated and a heatmap was drawn by Python. Gephi was used to visualize the multimorbidity network. Subgroup analysis was performed based on the clustering results and gender.
Results: This study included 54,829 individuals, with 30,872 females (56.3%). The average age was 75.9±7.1, and the prevalence of multimorbidity was 45.5%. The heatmap revealed the closest relationship between gout and osteoarthritis, with a correlation coefficient of 0.6. The cluster analysis revealed three multimorbidity patterns: the CAD-hypertension-cardiac failure cluster, the bronchical diseases-COPD-asthma cluster, the arrhymia-hyperlipidemia-osteoporosis cluster. The network analysis confirmed the strongest connection between gout and osteoarthritis, with a weight of 1.1. Subgroup analysis based on the clustering results indicated that within the arrhymia-hyperlipidemia-osteoporosis cluster, the relationship between hyperlipidemia and osteoporosis was the most tightly linked, with a weight of 0.1. In the bronchical diseases-COPD-asthma cluster, the connection between bronchial diseases and COPD was the closest, with a weight of 0.5. In the CAD-hypertension-cardiac failure cluster, the relationship between CAD and hypertension was the strongest, with a weight of 0.4. Gender-based subgroup analysis revealed that network density among females was higher at 0.83 compared to males at 0.78.
Conclusions: Multimorbidity is prevalent and females exhibited greater complexity in their multimorbidity patterns. These can facilitate clinicians in identifying core diseases and providing targeted interventions to lower the risks of multimorbidity.