Building Europe's First AI-Powered Childhood Obesity Prevention Network
Published in General & Internal Medicine and Public Health
The growing challenge of childhood obesity
The childhood obesity epidemic has a significant impact not only on children/adolescents and their families, but healthcare systems as a whole. What makes childhood obesity particularly challenging are its complex nature and its lifelong impact. Diet and exercise are not the only two factors involved; obesity management must rather consider a network of genetic, environmental, and behavioral factors; all interacting in ways that traditional research still struggles to fully understand. This complexity drives our pursuit towards advanced computational methodologies and a multidisciplinary approach. Moreover, by addressing childhood obesity, we're not just helping children today - we're potentially preventing decades of health challenges and improving the life quality of incoming generations.
From local solutions to a European network
Our initial research started from a key question: Why do behavioral interventions addressing childhood obesity work for some children but show little effect for others? This variation in treatment response suggested we are missing important pieces of the puzzle. The complexity of the topic led us to advance a further question: Could we use machine learning and a connected data network to identify the different obesity risks and better predict the effects of non-pharmacological interventions? Steming from this we defined the BIO-STREAMS project. Its objective is to deliver a digital biobank that connects genetic, environmental, and behavioral factors, helping uncover how these elements interact to shape each child's health. Initially, we will link clinical data from children suffering from overweight or obesity in seven clinical settings across six European countries, alongside data generated by children in seven European educational settings.
Building a system able to handle different types of data, originating from multiple countries, poses technical challenges. In addition, it is also important to ensure that this system is useful and usable in real-world healthcare settings. From the start, we have been applying Value Sensitive Design methodology, making sure our technical solutions would work well bad be valuable for everyone involved - from doctors and researchers to education professionals, families and children.
Innovation in healthcare technology
Our approach leverages artificial intelligence and machine learning to develop models capable of analyzing diverse data types, including genetic markers, lifestyle factors, clinical data and self-reports. At the same time, we prioritize transparency, ensuring stakeholders can understand and trust our methods. Beyond the medical community, we also consider the needs of schools and families to maximize the impact of these insights. Our initial dataset of 8,800 children provides an unprecedented opportunity to uncover patterns in obesity development, which will be further enriched with data from prospective studies in both clinical and educational settings.
A new approach to obesity prevention
What makes our work different is how the project embodies true collaborative design. We've not only brought together experts from different fields - clinicians, education professionals, data scientists, bioethicists, and healthcare economists - but we are also actively involving families, healthcare providers, and community representatives in our design process.
And we’re very excited about what’s coming next. The AI and machine learning models we will develop won't just remain in the lab; we've designed two prospective validation studies to test them in real-world settings. Thus, the translational potential of this research extends beyond retrospective analysis. The first will be implemented in clinical environments across our partner hospitals, while the second deploys a modified version in formal education settings. The parallel studies will help us to further understand how our digital interventions and biobank perform across clinical and non-clinical contexts, aiding healthcare providers, educators, and parents in identifying at-risk children and promoting healthier behavior in the community.
As we progress, we're not just collecting data - we're building and validating a foundation for a new direction in childhood obesity prevention. Our goal is that the BIO-STREAMS biobank and digital interventions, proven through real-world testing in both clinical and formal educational settings, will serve as a model for how healthcare systems can use artificial intelligence responsibly and effectively to address major public health challenges.
Funding
This work has been funded by the European Union Horizon Europe Research and Innovation Program project BIO-STREAMS (grant number 101080718), and the Swiss State Secretariat for Education, Research and Innovation (SERI). The content of this work does not reflect the official position of the European Union or any other institution. The information and views expressed are the sole responsibility of the authors.
Contacts
Follow the Topic
-
ISRCTN registry
A primary clinical trial registry recognised by WHO and ICMJE that accepts studies involving human subjects or populations with outcome measures assessing effects on human health and well-being, including studies in healthcare, social care, education, workplace safety and economic development.
Please sign in or register for FREE
If you are a registered user on Research Communities by Springer Nature, please sign in