The classification of diabetes and prediabetes by static glucose thresholds obscures the pathophysiological dysglycemia heterogeneity, primarily driven by insulin resistance (IR), beta-cell dysfunction, and incretin deficiency. This review demonstrates that continuous glucose monitoring and wearable technologies enable a paradigm shift towards non-invasive, dynamic metabolic phenotyping. We show evidence that machine learning models can leverage high-resolution glucose data from at-home, CGM-enabled oral glucose tolerance tests to accurately predict gold-standard measures of muscle IR and beta-cell function. This personalized characterization extends to real-world nutrition, where an individual's unique postprandial glycemic response (PPGR) to standardized meals, such as the relative glucose spike to potatoes versus grapes, could serve as a biomarker for their metabolic subtype. Moreover, integrating wearable data reveals that habitual diet, sleep, and physical activity patterns, particularly their timing, are uniquely associated with specific metabolic dysfunctions, informing precision lifestyle interventions. The efficacy of dietary mitigators in attenuating PPGR is also shown to be phenotype-dependent. Collectively, this evidence demonstrates that CGM can deconstruct the complexity of early dysglycemia into distinct, actionable subphenotypes. This approach moves beyond simple glycemic control, paving the way for targeted nutritional, behavioral, and pharmacological strategies tailored to an individual's core metabolic defects, thereby paving the way for a new era of precision diabetes prevention.

Diabetes is a chronic disease that poses a significant global health burden, and optimizing diabetes management requires multi-stakeholder collaboration. Large language models (LLMs) have shown promise in various healthcare scenarios, but their effectiveness across a diverse range of diabetes tasks remains unproven. In this study, we introduced a framework to train and validate diabetes-specific LLMs. We first developed a comprehensive data processing pipeline that includes data collection, filtering, augmentation and refinement. This approach contributes to creating a high-quality, diabetes-specific dataset, and several evaluation benchmarks entirely from scratch. Utilizing the collected training dataset, we fine-tuned a diabetes-specific LLM family that demonstrated state-of-the-art proficiency in understanding and processing various diabetes tasks compared to other LLMs. Furthermore, clinical studies showed the potential applications of our models in diabetes care, including providing personalized healthcare, assisting medical education, and streamlining clinical tasks. In conclusion, our study introduced a framework to develop and evaluate a diabetes-specific LLM family, and highlighted its potential to enhance clinical practice and provide personalized, data-driven support for diabetes support when facing different end users.

This Digital Health News industry roundup includes a new online course for young people to build skills for a future career in care, a milestone for digital-first healthcare-at-home company Cera and the integration of Ibex Medical Analytics' AI platform with Source BioScience's pathology network. A partnership between Babyl – a subsidiary of Babylon – and Novo Nordisk will help contribute to the expansion of diabetes awareness and care in Rwanda through community engagement and skills building using digital technology. Babyl's existing infrastructure and digital tech will be used to offer digital consultations to patients across Rwanda. Patients who then receive a confirmed diagnosis will be guided to the correct level of care by a doctor or nurse. This could include medication or a referral for further tests.

Since its inception as an academic discipline in 1956, Artificial Intelligence (AI) has enabled significant breakthroughs in the fields of science, healthcare, and transport. While the term'AI' may conjure up images of self-driving cars or sentient cyborgs, it's also transforming diabetes care and the way we identify, treat, and monitor disease. Here, I provide three examples of the impact that AI is having on diabetes care across the following key areas: 1) patient enablement and support, 2) data-driven approaches to disease prediction; and 3) clinician support. First up is the blossoming field of AI-driven continuous glucose monitoring devices. These devices are revolutionising the management of type-1 diabetes by providing automatic and real-time data of the rate of change and concentrations of blood glucose.

Conversational agents (CAs) represent an emerging research field in health information systems, where there are great potentials in empowering patients with timely information and natural language interfaces. Nevertheless, there have been limited attempts in establishing prescriptive knowledge on designing CAs in the healthcare domain in general, and diabetes care specifically. In this paper, we conducted a Design Science Research project and proposed three design principles for designing health-related CAs that embark on artificial intelligence (AI) to address the limitations of existing solutions. Further, we instantiated the proposed design and developed AMANDA - an AI-based multilingual CA in diabetes care with state-of-the-art technologies for natural-sounding localised accent. We employed mean opinion scores and system usability scale to evaluate AMANDA's speech quality and usability, respectively. This paper provides practitioners with a blueprint for designing CAs in diabetes care with concrete design guidelines that can be extended into other healthcare domains.

According to Søren Smed Østergaard, Vice President, Digital Health of Novo Nordisk, the most significant innovations in the diabetes space centered around hardware, artificial intelligence (AI) and data. He believes that having access to more accurate data on individual behavior and medication usage could positively impact people living with diabetes. "We know there is a huge discrepancy between how people should use medication and how they're using it," said Østergaard. "In 2003, the World Health Organisation (WHO) said improving medication adherence will have a more significant impact on the population's health than improvements to specific medical treatments. "Healthcare data today is often incomplete and too sparse to use for effective decision-making; we need to solve that first, but with this comes a plethora of ethical implications," said Østergaard. "People must have confidence that their data is being kept secure and used responsibly. Data sharing – creating a complete picture using data from different parties and devices – has the potential to revolutionize healthcare and outcomes, but robust data privacy policies must underpin it.

The effect that food has on blood glucose levels in people with Type 1 diabetes is well established. Less clear, however, is the role that stress, time of day, activity levels and other factors play in regulating blood glucose. To better understand these dynamics, UB researchers have launched a project that combines artificial intelligence (AI) with data gathered by continuous glucose monitoring tools. Ultimately, the goal is to better understand the relationship between insulin and blood glucosen, empowering people with Type 1 diabetes to better manage the condition and improve their quality of life. "We're developing new tools -- combining data collected from diabetes-monitoring tools with AI systems, as well as traditional time-series modeling approaches -- that could greatly improve how people manage their Type 1 diabetes," says the project's leader, Tarunraj Singh, professor of mechanical and aerospace engineering, School of Engineering and Applied Sciences.

The affect that food has on blood glucose levels in people with Type 1 diabetes is well established. Less clear, however, is the role that stress, time of day, activity levels and other factors play in regulating blood glucose. To better understand these dynamics, University at Buffalo researchers have launched a research project that combines artificial intelligence (AI) with data gathered by continuous glucose monitoring tools. Ultimately, the goal is to better understand the relationship between insulin and blood glucosen, empowering people with Type 1 diabetes to better manage the condition and improve their quality of life. "We're developing new tools -- combining data collected from diabetes monitoring tools with AI systems, as well as traditional time-series modeling approaches -- that could greatly improve how people manage their Type 1 diabetes," says the project's leader, Tarunraj Singh, PhD, professor of mechanical and aerospace engineering in the School of Engineering and Applied Sciences.

Machine learning has immense potential for improving diabetes care, particularly when used by diabetes care and education specialists, according to two presenters at the American Association of Diabetes Educators annual meeting. "We are in no way saying, and I in no way believe at all, that machine learning will replace diabetes education," Mark Heyman, PhD, CDE, vice president of clinical operations of medical technology company One Drop, said during the presentation. "But we really want to understand how it fits within the context of the role of the educator." With new technology making it easier than ever to record information about blood glucose levels, physical activity and meals, among other parameters, the amount of personal health information is exploding. Finding easy ways to put these data to use can be a challenge, which is what machine learning aims to address.

Type 2 diabetes mellitus (T2DM) is a chronic disease that often results in multiple complications. Risk prediction and profiling of T2DM complications is critical for healthcare professionals to design personalized treatment plans for patients in diabetes care for improved outcomes. In this paper, we study the risk of developing complications after the initial T2DM diagnosis from longitudinal patient records. We propose a novel multi-task learning approach to simultaneously model multiple complications where each task corresponds to the risk modeling of one complication. Specifically, the proposed method strategically captures the relationships (1) between the risks of multiple T2DM complications, (2) between the different risk factors, and (3) between the risk factor selection patterns. The method uses coefficient shrinkage to identify an informative subset of risk factors from high-dimensional data, and uses a hierarchical Bayesian framework to allow domain knowledge to be incorporated as priors. The proposed method is favorable for healthcare applications because in additional to improved prediction performance, relationships among the different risks and risk factors are also identified. Extensive experimental results on a large electronic medical claims database show that the proposed method outperforms state-of-the-art models by a significant margin. Furthermore, we show that the risk associations learned and the risk factors identified lead to meaningful clinical insights.