The Use of Smart Devices in Capturing Digital Biomarkers in Eating Disorders

Eating disorders are complex mental health conditions characterised by abnormal eating habits and distressing thoughts about body weight and shape, significantly impacting both mental and physical health. According to the International Classification of Diseases, 11th Revision, feeding and eating disorders include several subtypes such as anorexia nervosa (AN), bulimia nervosa (BN), and binge eating disorder (BED). These subtypes often involve patterns of restrictive or excessive food intake, leading to severe health issues and impairments in daily functioning.

Given the rising incidence and profound impacts of eating disorders in recent years, including high mortality rates and significant economic burdens, there is an urgent need for innovative management strategies and more efficient triage and monitoring systems for early intervention and home-based care.

For the care of individuals with eating disorders, comprehensive and continuous assessment of physical and psychiatric conditions is essential. As recommended by the UK's National Institute for Health and Care Excellence (NICE) and the Royal College of Psychiatrists, key clinical markers monitored, include weight loss, BMI, heart rate, blood pressure, temperature, hydration status, and muscular function. Monitoring these markers is crucial for early detection of medical complications, such as electrolyte imbalances, cardiac arrhythmias, and orthostatic hypotension, which pose serious health risks. Additionally, these markers help evaluate the disorder's severity, guide treatment adjustments, and ensure patient safety during recovery.

However, traditional methods used to assess these biomarkers are burdensome and time-consuming. With the advancement of technology, novel smart devices can efficiently detect traditional biomarkers such as heart rate and blood pressure, while also exploring potential novel measures of the disease. Therefore, this study aims to validate and explore the potential of these technologies to provide monitoring comparable to traditional methods and to integrate collected data to generate sophisticated insights into the health status of individuals with eating disorders.

According to previous research, the analysis of facial information, including static features and dynamic movements, combined with advanced algorithms and machine learning, can estimate body weight, BMI, parotid gland size, and skin condition. When voice pattern analysis is integrated with facial dynamics during the video diary entry phase and the image response task, where participants share their thoughts on specific images, it is expected to further assess physical and psychological states, particularly when discussing sensitive topics or images, such as high-calorie foods. These estimations can be used to interpret the health status of individuals with eating disorders. Additionally, photoplethysmography (PPG) using smart devices can detect subtle changes in the colour spectrum induced by blood volume dynamics in facial and fingertip areas, allowing for the estimation of heart rate (HR), blood pressure (BP), respiratory rate (RR), blood oxygen level, blood glucose, and body temperature. This technology, which has been validated in healthy subjects, shows significant potential for application in patients with eating disorders, who are prone to cardiovascular and respiratory issues due to physiological stress and nutritional imbalances. This approach provides essential insights into their physical health status, particularly given the significant impairments in muscle strength often observed in individuals with eating disorders.