Loneliness, Insomnia and Poor Mental Health Linked to 35% Higher Diabetes Risk
A groundbreaking new study has revealed that common psychological and behavioural factors significantly increase the risk of developing type 2 diabetes. Research led by Anglia Ruskin University (ARU) found that loneliness, insomnia, and mental health problems such as depression and anxiety each independently elevate diabetes risk by an estimated 35 percentage points.
Chronic Stress Response Drives Metabolic Disruption
The study, which analysed lifestyle and health data from 19,774 adults in the UK Biobank tracked for up to 17 years, suggests these links are likely driven by the body's response to chronic stress. Researchers used artificial intelligence (AI) to predict and simulate disease development, finding that psychological factors "likely reflect" well-established stress responses including inflammation, impaired blood sugar regulation, and overproduction of the stress hormone cortisol.
"These effects underscore that psychosocial distress is not just a mental challenge, but a potent metabolic disruptor with real and measurable health consequences," the researchers stated. When all three factors—loneliness, insomnia, and poor mental health—were combined, the estimated diabetes risk increased dramatically to 78 percentage points.
Overlooked Factors Provide Meaningful Risk Signals
Dr Mahreen Kiran, lead author and postgraduate researcher at ARU, emphasized the importance of including these often-ignored variables in health datasets. "Behavioural and psychosocial factors such as loneliness, sleep disruption and mental health history are frequently overlooked in traditional risk assessment," she explained. "Yet our research demonstrates they provide meaningful signals about future disease risk that should not be ignored."
The study also identified a concerning link between these psychological stressors and dietary patterns. People experiencing loneliness, insomnia, and mental health issues were more likely to consume diets high in salty, sugary cereals and processed meats—dietary choices that independently contribute to increased diabetes risk.
Current Risk Models Over-Simplify Complex Disease
Professor Barbara Pierscionek, deputy dean for research and innovation at ARU's Faculty of Health, Medicine and Social Care, criticized current diabetes prediction models for their oversimplification. "Type 2 diabetes is a rising global health concern heavily influenced by lifestyle," she noted. "However, existing risk prediction models rely primarily on BMI, age and blood pressure, overlooking the complex interconnected behavioural and emotional factors that precede and shape disease onset."
The research, published in Frontiers in Digital Health, utilized an innovative digital twin model system that employs AI to replicate individual health profiles and simulate disease progression. This approach enables researchers to test "what-if" scenarios and potentially tailor interventions to specific patient needs.
Digital Twin Models Offer Cost-Effective Solutions
Professor Pierscionek highlighted both the potential and limitations of digital twin technology for diabetes management. "Digital twin model systems present a viable cost-effective way of diagnosis, testing and treatment for numerous conditions," she said. "However, most existing models depend on real-time data from wearable devices, creating barriers for settings lacking technical infrastructure or underserved communities struggling with costs."
The findings come at a critical time for UK healthcare, with approximately 4.6 million people currently diagnosed with diabetes and estimates suggesting an additional 1.3 million may be living with undiagnosed type 2 diabetes. The research provides compelling evidence that healthcare professionals must consider psychological wellbeing alongside traditional physical health indicators when assessing diabetes risk and developing prevention strategies.
