A Dynamic Approach to Understanding the Impact of Health Behaviours on Mental Health in Diabetes

Despite advancements in diabetes clinical care, the psychological burden of living with diabetes remains a challenge. Health behaviours, such as physical activity, sleep, diet, smoking, and alcohol consumption have been associated with mental health outcomes in the general population longitudinally, and among those with diabetes cross-sectionally. Clarifying the role of health behaviours in changes in mental health and related psychological factors over time in people with diabetes could help us understand, prevent, and treat mental health comorbidity in people with diabetes. Furthermore, traditional research and clinical assessments have relied on retrospective self-reports, summary scores on mental health assessments, and infrequent measurements, potentially missing crucial micro-level changes and symptom interactions. Health behaviours are dynamic and context dependent and mental health may be driven by complex, dynamic systems of sub-clinical mental states. Therefore, we also need assessments and techniques that are dynamic and that take account of symptom-level and daily experience-level interactions. The overarching aim of this thesis was to investigate the longitudinal impact of health behaviours on mental health in people with diabetes. To do so, it adopted a dynamic approach, focusing on micro-level connections between health behaviours and sub-clinical mental states, which encompasses both momentary experiences of mood and individual mental health symptoms. This thesis consists of 4 studies: (1) a scoping review synthesising existing evidence on the relationship between health behaviours and mental health and related psychological factors in people with diabetes; (2) an assessment of the feasibility, acceptability, and user-experience of a digital phenotyping study; (3) a novel digital phenotyping study examining the impact of daily health behaviours on daily mood; and (4) a network analysis of health-risk behaviours and individual depressive symptoms using multinational secondary data. Results from Studies 1, 3 and 4 indicate that health-promoting behaviours, especially exercise and movement, positively influence mental health and related psychological factors, (including sub-clinical mental states) in people with diabetes. The evidence for the impact of health-risk behaviours, aside from physical inactivity, was weaker. This pattern was also observed for people without diabetes, however, micro-level examination indicated nuanced differences in how health behaviours and sub-clinical mental states interact between people with and without diabetes. In particular, physical (in)activity and sleep may have different micro-level relationships with sub-clinical mental states for people with and without diabetes, and momentary anger may be differentially impacted by physical activity for people with diabetes compared to those without. Finally, Study 2 found that the benefits of a key micro-level methodology, digital phenotyping, may be more apparent to people with type 2 diabetes than those without. The insights obtained from this thesis contribute to a more comprehensive and nuanced understanding of the connections between health behaviours and mental health, particularly sub-clinical mental states, for people with diabetes and could inform the development of more precise interventions, ultimately leading to improved overall well-being and quality of life for people with diabetes.