Characterization of diseases in temporal comorbidity networks

Comorbidity networks, or disease-disease networks, represent relationships between diseases and can reveal non-random disease associations and multimorbidity structure. These patterns reflect shared pathophysiology, diagnostic pathways, and risk of adverse outcomes. Using large-scale hospital records, these networks make it possible to study how disease relationships change across the life course. Yet how age-related changes in network structure relate to disease prevalence and mortality remains understudied. Here we used publicly available comorbidity networks extracted from a comprehensive dataset of 45 million Austrian hospital stays from 1997 to 2014, covering 8.9 million patients. Networks become substantially denser with age, growing from 111 nodes (degree 5.01) in childhood to 384 nodes (degree 21.78) in elderly females. We identified diseases that showed disproportionate connectivity relative to prevalence, including iron deficiency anemia (D50) in children and nicotine dependence (F17) and dyslipidaemia (E78) in adults. By linking topology to in-hospital mortality, we identified high-mortality and central nodes, such as cancers, liver cirrhosis (K74), chronic kidney disease (N18), COPD (J44), and heart failure (I50). We also identified low-to-high mortality central edges, including cardiometabolic risk factors that connect to fatal cardiovascular events and aggressive cancers. These findings underscore the importance of accounting for age- and sex-specific network structure when interpreting comorbidity patterns. They point to a small set of disproportionately connected, high-mortality “bridging” conditions that may warrant further investigation. Future studies could examine the biological and other underlying factors that may contribute to the observed network associations.