Objective To investigate the prevalence and common comorbidity patterns of chronic diseases among elderly inpatients in Shenzhen city, Guangdong province, from 2017 to 2022, and to provide scientific evidence for the development of comorbidity management and prevention strategies.

Methods Data from the discharge summaries of 410 814 inpatients aged 65 years and older from 142 primary and higher-level hospitals in Shenzhen city between January 1, 2017, and December 31, 2022, were collected. Frequency combination analysis was used to identify common chronic disease comorbidity combinations, and association rule analysis was used to identify common non-random disease combinations.

Results Among the 410 814 elderly inpatients in Shenzhen city from 2017 to 2022, 387 020 (94.21%) had chronic disease comorbidities. Hypertension was the most prevalent chronic disease (59.14%), accounting for 97.79% of the comorbidity composition. The most common two- and three-disease combinations were “hypertension + stroke” (106 005 cases, 25.80%) and “hypertension + peripheral artery disease + stroke” (55 343 cases, 13.47%), respectively. The top 20 two- and three-disease combinations mainly consisted of hypertension, stroke, ischemic heart disease, heart failure, peripheral artery disease, dyslipidemia, and type 2 diabetes, with cardiovascular diseases predominating, followed by metabolic diseases. Association rule analysis showed that among the top 20 rules, the rule “(refractive errors and accommodation disorders) => (cataract)” had the highest lift value of 6.128. The remaining association rules all involved cardiovascular and cerebrovascular diseases, with the rule “(chronic ischemic heart disease, hypertension, peripheral artery disease) => (heart failure)” having a lift value of 3.082.

Conclusions The prevalence of chronic disease comorbidities among elderly inpatients in Shenzhen city is high. Cardiovascular diseases dominate the two- and three-disease combinations. Cataracts and cardiovascular and cerebrovascular diseases are the main non-random comorbidity association patterns.