Investigation of multimorbidity patterns and association rules in patients with type 2 diabetes mellitus using...

The issue of multimorbidity in patients with type 2 diabetes mellitus (T2DM) is extremely serious. However, the pattern of multimorbidity, including typical complications, remains unclear. This study aims to explore the current status and influencing factors of multimorbidity in T2DM, with a focus on mining frequent disease combination patterns and strong association rules. Data on 26 diseases were extracted from the electronic medical records of 5,838 hospitalized patients with type 2 diabetes. The chi-square test, Cochran-Armitage trend test, and logistic regression were used for the analysis of influential factors. Association rule mining was employed to explore frequent disease combinations and association rules across the entire population and subgroups stratified by gender, age, and BMI. Network graphs were used to visualize binary comorbidity relationships. Gender-specific differences in disease prevalence were found for 18 of the 26 diseases included in this study. The prevalence of multimorbidity was 97.8%, and it increased with age, with a higher prevalence in males (P < 0.05). The identified frequent disease combination patterns mainly centered around typical complications of T2DM. The most frequent binary comorbidity pattern was diabetic peripheral neuropathy (DPN) + diabetic peripheral vascular disease (DPVD) (support: 74.1%), which is a novel finding in the relationship between DPN and DPVD. The primary association rule identified was {DPVD + diabetic nephropathy (DN)}→{Hypertension}. Disease combination patterns and association rules varied across gender, age, and BMI. Comorbidity relationships became more complex in the middle and older age groups, as well as in the overweight and obese groups. The findings of this study can be used to guide clinicians in the prevention and treatment of multimorbidity in T2DM and provide possible directions for researchers to further investigate the causes and mechanisms.