Data-driven identification of complex disease phenotypes

Abstract

Disease interaction in multimorbid patients is relevant to treatment and prognosis, yet poorly understood. In the present work, we combine approaches from network science, machine learning and computational phenotyping to assess interactions between two or more diseases in a transparent way across the full diagnostic spectrum. We demonstrate that health states of hospitalized patients can be better characterized by including higher-order features capturing interactions between more than two diseases. We identify a meaningful set of higher-order diagnosis features that account for synergistic disease interactions in a population-wide ( N = 9 M) medical claims dataset. We construct a generalized disease network where (higher-order) diagnosis features are linked if they predict similar diagnoses across the whole diagnostic spectrum. The fact that specific diagnoses are generally represented multiple times in the network allows for the identification of putatively different disease phenotypes that may reflect different disease aetiologies. At the example of obesity, we demonstrate the purely data-driven detection of two complex phenotypes of obesity. As indicated by a matched comparison between patients having these phenotypes, we show that these phenotypes show specific characteristics of what has been controversially discussed in the medical literature as metabolically healthy and unhealthy obesity, respectively. The findings also suggest that metabolically healthy patients show some progression towards more unhealthy obesity over time, a finding that is consistent with longitudinal studies indicating a transient nature of metabolically healthy obesity. The disease network is available for exploration at https://disease.network/.

Exploring multimorbidity profiles in middle-aged inpatients: a network-based comparative study of China and the United Kingdom

“…The current research on multimorbidity is riddled with significant knowledge gaps, despite numerous efforts. Existing studies are largely limited to developed countries, with only limited research from developing countries [ 10 – 13 ]. Besides, the current research has focused on older adults [ 14 – 16 ] but not the younger middle-aged population.…”

“…Besides, the current research has focused on older adults [ 14 – 16 ] but not the younger middle-aged population. More importantly, most existing studies have focused on a small group of chronic diseases, with limited exploration of multimorbidity on a wider spectrum of diseases, except only a handful of studies in Southwest China [ 17 ], Denmark [ 11 ], America [ 12 ], and Austria [ 13 ]. To date, no studies comprehensively compared comorbidity patterns at a population level between developed and developing countries.…”

“…Complex disease interactions significantly influence diagnosis and treatment and pivotal inpatient care. Untangling these less-understood relationships is crucial [ 13 , 18 ]. Our study employed network analysis, an innovative tool revealing intricate interconnections among diseases [ 10 – 13 ].…”

“…Untangling these less-understood relationships is crucial [ 13 , 18 ]. Our study employed network analysis, an innovative tool revealing intricate interconnections among diseases [ 10 – 13 ]. We used this approach to construct multimorbidity networks in China and the UK, identifying hub diseases with numerous comorbidity patterns.…”

“…The diagnosis of diseases is according to the International Statistical Classification of Diseases and Related Health Problems 10th revision (ICD-10) [ 19 ]. We included 14 systematic chapters on diseases and used sophisticated network analysis to develop multimorbidity networks, compare comorbidity patterns, and identify hub diseases between China and the UK [ 13 ]. Our findings are pivotal in global healthcare by highlighting disease disparities across countries with different economic development statuses.…”