Knowledge Gaps in Behavioural Science Relating to Type 2 Diabetes Prevention: A Narrative Review

The review addresses the growing global and UK burden of type 2 diabetes and emphasizes that key risk factors—poor diet and physical inactivity—are behavioural and modifiable. Behavioural interventions can reduce incidence among people at elevated risk and are often cost‑effective at both population and individual levels. The review aims to: (a) identify how whole‑population approaches can be made more effective and equitable over the long term; (b) determine how high‑risk Diabetes Prevention Programmes (DPPs) can be optimized for engagement, maintenance, mechanisms, and delivery; and (c) assess whether stress and other emotional determinants warrant greater consideration in prevention. The purpose is to clarify current knowledge and highlight gaps to guide future behavioural science research in diabetes prevention.

The narrative review spans population‑level and high‑risk approaches. It appraises evidence that behaviour change for diabetes prevention often mirrors generic weight management strategies and that disease‑specific diet or activity recommendations are not currently supported by evidence. It examines equity and effectiveness of micro‑environmental (choice architecture) interventions across socioeconomic groups, noting inconsistent and sparse moderation evidence and calling for large studies and meta‑analyses. Long‑term impacts of population measures are under‑evaluated; supermarket interventions are typically small scale and short duration. For DPPs, the review synthesizes evidence on reach and engagement (e.g., NHS‑DPP’s >500,000 referrals in four years), disparities by age and socioeconomic position, and potential of digital delivery with human coaching to enhance engagement. It examines maintenance of behaviour change and weight loss, highlighting typical weight regain and limited long‑term physical activity effects (e.g., PROPELS). Mechanistic understanding is constrained by under‑reporting of intervention content and limited fidelity assessments; it advocates for logic models, TIDieR, and BCT Taxonomy reporting. It reviews which BCTs are effective (self‑regulation techniques) and gaps regarding interactions, timing, and dose, especially across older and diverse populations. It discusses modes of BCT delivery (group vs one‑to‑one; digital with coaching) and explores innovative approaches including third‑wave CBT (e.g., ACT), cognitive training, leveraging social context, and just‑in‑time adaptive interventions (JITAIs). The mental health section summarizes evidence linking depression, anxiety, eating disorders, and severe mental illness to elevated diabetes risk and poorer DPP outcomes, noting the need for integrated mental health support within prevention programmes.

Narrative review. No systematic search or formal consensus procedures were undertaken. Authors, invited by the Diabetes UK Research Steering Group on Prevention and Management of Diabetes based on expertise, developed recommendations through discussions, email correspondence, and iterative drafting. Disagreements were minor and resolved through iteration. Emphasis is on identifying knowledge gaps and research priorities rather than providing a comprehensive systematic synthesis.

- Disease‑specific behavioural prescriptions: Current evidence does not support diet or physical activity guidance unique to type 2 diabetes prevention; general population recommendations suffice, with weight loss being the primary mechanism. Combining prevention messaging across multiple diseases may enhance engagement. - Population vs high‑risk approaches: Upstream population‑level measures (fiscal, media, micro‑environmental nudges) are necessary complements to high‑risk DPPs. Equity of impact for micro‑environmental interventions remains under‑tested; moderation by socioeconomic position is unclear with mixed review findings. - Long‑term and scalable population interventions: Many environmental interventions are short‑term and small‑scale; there is a need for sustained, scalable implementations (e.g., supermarket interventions) to evaluate durability and moderators and to improve economic modelling. - DPP engagement: NHS‑DPP referred >500,000 individuals in four years, but participation skews older and under‑represents lower SEP groups. Digital DPPs can reach younger or time‑constrained individuals; human coaching in digital programmes increases engagement with goal setting and self‑monitoring. - Maintenance: Small sustained weight losses (≈1.5–2 kg) meaningfully reduce incidence, yet weight regain is common and glycaemia rises with regain. Long‑term effects of digital programmes beyond 12 months are under‑studied. PROPELS showed initial increases in steps at 12 months (≈+532/day) that were not sustained at 48 months. More diverse maintenance strategies (e.g., ongoing support, stepped care, rescue plans) require testing. - Mechanisms and fidelity: Under‑reporting of intervention content hampers identifying active ingredients. Greater use of logic models, TIDieR, BCT Taxonomy, and fidelity assessment is needed. Self‑regulation BCTs are supported, but evidence on BCT combinations, timing, dose, and effectiveness across demographics is limited. Older or less health‑literate groups may require additional support for understanding/enacting BCTs. - Delivery modes: Some BCTs suit digital self‑delivery (e.g., self‑monitoring), while others (goal setting, feedback, problem solving) benefit from human coaching. Comparative analyses across providers could identify effective delivery modalities. - New approaches: Third‑wave CBTs (e.g., ACT) show promise for weight and glycaemic outcomes; cognitive training may aid selected individuals but faces engagement challenges; leveraging social networks can both help and hinder behaviour change; JITAIs are promising but untested specifically for diabetes prevention. - Mental health: Depression, anxiety, eating disorders, and severe mental illness increase diabetes risk and can impede adherence and outcomes in DPPs. Tailored support and explicit integration of mental health within prevention are needed to reduce distress and stigma while motivating engagement. - Box 1 research priorities: Calls for long‑term, scalable population interventions; equity assessments; evaluation of disease‑agnostic vs disease‑specific approaches; optimization of engagement, maintenance, mechanisms (BCTs, timing/dose, fidelity), delivery modes (including coaching), incorporation of psychological flexibility and social context, and development/testing of JITAIs; and explicit attention to mental health within prevention.

The review frames behavioural science as central to preventing type 2 diabetes and argues that optimizing both population‑level and high‑risk strategies is essential for maximal and equitable impact. By synthesizing evidence on engagement, maintenance, and mechanisms, it highlights that current DPPs, while effective, can be refined to improve reach (particularly among lower SEP and minority ethnic groups), sustain behaviour change, and reduce health inequalities. It underscores that self‑regulation techniques, though foundational, may need augmented support and alternative or complementary strategies (e.g., ACT, social network leverage, JITAIs) to address diverse needs and contexts. Integrating mental health support is pivotal to improving adherence, reducing distress, and enhancing overall effectiveness. Enhanced reporting standards and fidelity measures are necessary to identify and scale active ingredients, inform tailoring by demographics, and guide cost‑effective implementation.

Behavioural science has a clear role in preventing type 2 diabetes. Existing DPPs, adapted from protocols developed decades ago, are effective and scalable but require optimization to improve inclusivity, long‑term maintenance, and cost‑effectiveness. Future research should prioritize: sustained, scalable population interventions with equity evaluation; comparative evaluation of disease‑agnostic vs disease‑specific and tailored approaches; improved engagement through human‑supported digital delivery and alternative programme formats; deeper mechanistic insight via better reporting and fidelity; novel strategies enhancing psychological flexibility, leveraging social context, and deploying JITAIs; and explicit incorporation of mental health support. Advances here will benefit both prevention and management of diabetes.

As a narrative review, it did not undertake systematic literature searches, structured quality appraisal, or formal consensus, which may introduce selection and interpretation biases. Evidence gaps persist regarding equity impacts of environmental interventions, long‑term effectiveness of digital and maintenance strategies, fidelity and active ingredients of DPPs, and the generalizability of BCT findings to older and diverse populations. Limited reporting in primary studies constrains robust mechanistic conclusions.