Abstract
BACKGROUND: As populations are aging globally and healthcare systems are transitioning from being centred around single disease to patient-centred approach, precise and flexible identification of the multimorbidity patterns allows sound measures to prevent adverse health outcomes and better allocate healthcare resources. METHODS: Combining probabilistic approach of graphical model and intuitive visibility of network analysis, we use administrative health data of individuals aged 50 and above residing in Emilia-Romagna region (northern Italy) in 2011 and followed up to 2019 (N = 1,010,571) to investigate multimorbidity patterns and their impact across time. RESULTS: Four consistent multimorbidity patterns are identified across sex, across age groups (50–59, 60–69, 70–79, 80 +), and across time-points (2011, 2016 and 2019), which consists of cardiovascular, neuropsychiatric, respiratory-digestive and metabolic-pain pattern. This finding suggests plausible existence of stable population multimorbidity structure for public health monitoring. We also bring evidence of the better performance of multimorbidity patterns in mortality modelling compared to traditional methods. While neuropsychiatric pattern is the leading group causing mortality at older ages, higher education and living in suburban areas provide consistent protective effect against mortality, unplanned hospitalization and length of stay for all sexes and age groups. We focus on the gatekeeper diseases as potential targets for intervention. They are not necessarily the diseases with highest prevalence, but we demonstrate that an early detection of these diseases can contribute to improving three different health outcomes at older ages. CONCLUSION: For each sex and age group above 50, we provide systematically the network of chronic diseases, the corresponding multimorbidity patterns, and the estimated impacts of multimorbidity, and thus a comprehensive understanding of multimorbidity at older ages. The observed consistent multimorbidity structure at population level and the identification of target diseases for intervention that goes beyond the prevalence-based approach can open new doors for more directions shaping public health policy. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s13690-026-01886-3.