Pilot study of a ketogenic diet in bipolar disorder: a process evaluation

Background: Bipolar disorder is a serious mental illness, which requires new strategies for prevention and management. Recent evidence suggests that a ketogenic diet may be an effective intervention. This research aimed to explore the feasibility and acceptability of a ketogenic diet intervention for bipolar disorder, fidelity to its behavioural components and the experiences of the participants and research clinicians involved.

Methods: A mixed-methods process evaluation was conducted. Semi-structured telephone interviews were carried out with 15 participants 1-2 months after completing a 6-8 week modified ketogenic diet intervention, and 4 research clinicians from the study team following the completion of data collection. Data were thematically analysed. Fidelity checklists completed by research dietitians were analysed using descriptive count and percentage statistics. Findings are reported post-hoc, following the analysis and publication of the main pilot study findings. Reporting was guided by the COREQ checklist.

Results: Five themes were identified in the qualitative data: (1) 'Encouraging entry and supporting exit' (e.g. recognising and managing participants' varied motives and expectations, including around weight loss and symptom alleviation); (2) 'Challenging but potentially transformational,' which reflects that while it can be difficult to initiate and maintain a ketogenic diet day-to-day, many participants perceived physical and psychological benefits (e.g. significant weight loss, mood stability and enhanced ability to focus); (3) 'Intervention facilitators,' including a range of behavioural (e.g. goal setting), social (e.g. family and dietitians) and technological (e.g. apps for monitoring) support mechanisms; (4) 'Intervention barriers' (e.g. dietary preferences, concerns about the diet and its impact, the testing burden and capacity of the delivery team); and (5) 'The wider context' (e.g. the cost of living and sociocultural expectations) was a crucial factor explaining differential experiences. Overall, descriptive analyses indicated moderate-to-good fidelity to the behaviour change components of the study.

Conclusion: We provide novel insight into the experiences of people living with bipolar disorder initiating and following a ketogenic diet, as well as those of research clinicians who support the intervention. Future trials may benefit from increased clinical research capacity, better-defined entry and exit routes, additional interpersonal support, and greater understanding of how social and societal factors impact participation.

Trial registration: Study registration number: ISRCTN61613198 (02/03/22).