Detecting early warning signs and disease transition in alcohol addiction
A group of e:Med researchers from the SysMedAlcoholism consortium and their collaborators at the University of Tokyo have revealed for the first time evidence of a dynamic transition in health state at disease onset using rats.
The idea of so-called dynamical disease that regards disease onset as the dynamic transition of the body's internal state and the study of such disease dynamics have been drawing attention for their potential to inform the prediction of transitions into or between disease states. However, these ideas have not yet come to fruition or translated into clinical applications due to the lack of appropriate data and corresponding analytical techniques for making such predictions.
The research group led by Professor Rainer Spanagel at the ZI in Mannheim used their well-established alcohol deprivation effect (ADE) model, in which rats were allowed to consume alcohol freely for an initial period, followed by deprivation and subsequent reintroduction of the substance, to induce relapse-like excessive drinking, as an example of disease onset.
Over a period of 14 weeks (8 baseline, 2 deprivation, 4 reintroduction), the researchers acquired continuous and high-resolution intensive longitudinal data (ILD) of drinking behavior and locomotor activity, and analyzed the data using a multiscale computational approach. They found that transitions into addictive drinking behavior were preceded by predictive "early warning signals," such as unstable drinking patterns and instability in locomotor circadian (approximately 24-hour cycle) rhythms and a resultant increase in low-frequency ultradian rhythms (a few to several hour diurnal cycles) during the first week of the deprivation period.
The current findings are particularly relevant today, in which rapid developments in wearable and mobile biomedical sensing technology allow us to gather massive amounts of such biomedical intensive longitudinal data. The analytical framework developed in the present study has the potential of contributing to our understanding of disease onset in humans and forecasting changes into different stages of disease, and can be directly translated to the clinical arena through the appropriate use of these technologies for disease prevention.
"We believe that our approach provides a blueprint for biomedical and biophysical scientists who want to detect early warning signs and disease trajectories” says Rainer Spanagel. He continues, "Our new analytical framework for intensive longitudinal data will be of great help for many e-health applications."
For any correspondence please contact firstname.lastname@example.org
Jerome Clifford Foo, Hamid Reza Noori, Ikuhiro Yamaguchi, Valentina Vengeliene, Alejandro Cosa-Linan, Toru Nakamura, Kenji Morita, Rainer Spanagel, Yoshiharu Yamamoto , "Dynamical state transitions into addictive behavior and their early-warning signals ", Proc Biol Sci. 2017 pii: 20170882. doi: 10.1098/rspb.2017.0882