How is the NHS Low-Calorie Diet Programme expected to produce behavioural change to support diabetes remission: An examination of underpinning theory

Tamla S Evans,1 Rhiannon E Hawkes,2 Chris Keyworth,3 LIsa Newson,4 Duncan Radley,5 Andrew J Hill,6 Jamie Matu,1 Louisa J Ells1

1 Obesity Institute and School of Health, Leeds Beckett University, Leeds, UK
2 Manchester Centre for Health Psychology, Division of Psychology and Mental Health, University of Manchester, Manchester, UK
3 School of Psychology, Faculty of Medicine and Health, University of Leeds, Leeds, UK
4 School of Psychology, Faculty of Health, Liverpool John Moores University, Liverpool, UK
5 Obesity Institute and Carnegie School of Sport, Leeds Beckett University, Leeds, UK
6 Institute of Health Sciences, Faculty of Medicine and Health, University of Leeds, Leeds, UK

Address for correspondence: Miss Tamla S Evans
School of Health, Leeds Beckett University CL615, Calverley Building, City Campus, Leeds, LS1 3HE, UK


Background: In 2020, the National Health Service Low-Calorie Diet Programme (NHS-LCD) was launched, piloting a total diet (TDR) replacement intervention with behaviour change support for people living with Type 2 Diabetes (T2D) and excess weight. Four independent service providers were commissioned to design and deliver theoretically grounded programmes in localities across England.

Aims: 1) to develop a logic model detailing how the NHS-LCD programme is expected to produce changes in health behaviour, and (2) to analyse and evaluate the use of behaviour change theory in providers’ NHS-LCD Programme designs.

Methods: A documentary review was conducted. Information was extracted from the NHS-LCD service specification documents on how the programme expected to produce outcomes. The Theory Coding Scheme (TCS) was used to analyse theory use in providers’ programme design documents.

Results: The NHS-LCD logic model included techniques aimed at enhancing positive outcome expectations of programme participation and beliefs about social approval of behaviour change, to facilitate programme uptake and behaviour change intentions. This was followed by techniques aimed at shaping knowledge and enhancing the ability of participants to self-regulate their health behaviours, alongside a supportive social environment and person-centred approach.

Application and type of behaviour change theory within service providers’ programme designs varied. One provider explicitly linked theory to programme content; two providers linked 63% and 70% of intervention techniques to theory; and there was limited underpinning theory identified in the programme design documents for one of the providers.

Conclusion: The nature and extent of theory use underpinning the NHS-LCD varied greatly amongst service providers, with some but not all intervention techniques explicitly linked to theory. How this relates to outcomes across providers should be evaluated. It is recommended that explicit theory use in programme design and evidence of its implementation becomes a requirement of future NHS commissioning processes.

Br J Diabetes 2022;22:20-29

Key words: theory, intervention design, logic model, behaviour change, type 2 diabetes, diabetes remission, low-calorie diet, total diet replacement.


It is considered best practice for complex interventions to be theoretically grounded throughout the process of design, evaluation and implementation.1-3 Programme theory is a theory or model that explicitly describes how an intervention’s components expect to produce change and achieve intended outcomes.4 There are numerous benefits of theory use, including informing what psychosocial constructs should be targeted, selecting appropriate intervention techniques (sometimes referred to as an intervention’s active ingredients) to target these constructs, and deciding which research questions and measures are appropriate as part of a process evaluation.5 Through this, mechanisms of action can be understood (that is, how a programme is expected to achieve the desired outcomes), and intervention theory and design can be refined in light of findings. This has further benefits for intervention adaptation and scale-up, as in the absence of programme theory it is difficult to decipher which key mechanisms and techniques of interventions are necessary to retain for outcomes to be produced in other settings.1,6

Theory, therefore, facilitates the intervention development process from early conceptualisation and planning through to scale-up and adaptation through a structured framework. Health promotion interventions across a range of behaviours are argued to be more effective when a theory or theoretical constructs have been explicitly described.7 Since behaviour change is central to diabetes care, explicit use of underpinning behavioural science theory is argued to guide intervention developers and healthcare professionals in delivering evidence-based behaviour change support to optimise diabetes self-management.8

Despite these benefits, there is often a lack of explicit application of theory to intervention design and evaluation.5 There is often a misconception that an intervention is theory-based, leading to a growing concern that intervention developers are not using theory throughout the intervention development process. For example, one review of theory use in health promotion interventions found 69% of studies described an underpinning theoretical framework. However only 18% evidenced application, few (44%) tested changes in theoretical constructs, and 9% refined theory following evaluation.7,9 A meta-review found that 56% of physical activity and healthy eating interventions reported a theoretical basis, whilst only 10% of these reported links between all behaviour change techniques (BCTs) and theoretical constructs.10 Similar findings have been reported within the context of diabetes self-management programmes.8 This suggests that although the majority of intervention developers report theory use initially, evidence of theory being ‘an integral part of a rigorous scientific process’ is limited.5

One way in which intervention developers can articulate programme theory is through the construction of a logic model. Logic models provide a visual representation that maps out the theorised mechanisms of action and change, and can be a useful way of illustrating aspects of multiple theories that have informed an intervention’s design. In other words, a logic model is a process which represents the theory of how an intervention produces its outcomes, and typically considers: the intervention’s purpose (what motivates the need for change?); context (what is the climate in which change will take place?); inputs (what raw materials will be used to conduct the effort or initiative?); activities (what will the intervention do with its resources to direct the course of change?); outputs (what evidence is there that the activities were performed as planned?); and outcomes (what kinds of changes came about as a direct or indirect effect of the activities?).

A logic model creates a shared understanding of the programme assumptions and relationships between intervention components amongst the stakeholders and multidisciplinary teams involved in its development or implementation.11 Logic models are a useful guide for intervention design and evaluation as they outline the necessary techniques (e.g. BCTs) and the constructs that require assessment as part of a process evaluation. This is particularly useful for teams who were not involved in an intervention’s development who subsequently evaluate or adapt an intervention. Without a clear underpinning programme theory, consequences might include a loss in fidelity (i.e. the extent to which a programme is implemented as intended).12 Logic modelling has been used as a helpful tool across diabetes services; for example foot care,13 retinal screening,14 and the national diabetes prevention programme evaluation.12

In 2020, the National Health Service Low-Calorie Diet Programme (NHS-LCD) was launched.15 This is a weight loss intervention for patients living with comorbid T2D and overweight or obesity. Four independent weight management service providers were commissioned by NHS England (NHSE) to deliver theoretically grounded behaviour change support to achieve weight loss and improve T2D through a total diet replacement (TDR) approach. Patients were prescribed soups, bars and shakes providing ≤900kcal/d followed by a period of food reintroduction and weight maintenance support. The 52-week programme was initially piloted across 10 socioculturally diverse areas of England and expanded to a further 11 areas in 2022.

NHSE produced a service specification detailing the intervention contents that should be included in the NHS-LCD programme. This was based on new knowledge about the pathophysiology of T2D,16-18 and evidence demonstrating low-calorie TDR approaches to be the most effective dietary intervention for achieving T2D remission (HbA1c <48mmol/mol), followed by weight maintenance support to sustain outcomes.19-21 Within the service specification, commissioners stipulated that the programme must be theoretically grounded in behaviour change support.22 Previous research has evaluated the service delivery fidelity of the NHS Diabetes Prevention Programme (NHS-DPP), another behaviour change programme commissioned by NHSE and delivered by independent service providers that was aimed at adults in England who are at risk of developing T2D.12 Evaluators of this programme found explicit theory use in providers’ programme designs to be lacking. Importantly, none of the NHS-DPP providers had included a logic model in their programme designs to describe how their interventions were expected to work.12,23

The recently updated Medical Research Council (MRC) framework recommends that those identifying an intervention (i.e. those evaluating an intervention who were not involved in programme development) should uncover the underpinning programme theory where one has not been previously established, and should refine it during successive evaluation phases even if roll-out has begun.1 Furthermore, evaluators of the NHS-LCD have been commissioned to conduct a realist-informed evaluation to identify what works for whom, in what contexts and why [NIHR132075]. Without a logic model detailing how an intervention expects to produce behavioural changes and health outcomes, evaluators cannot identify the underpinning mechanisms and constructs that need to be evaluated in a process evaluation, thus inhibiting a theory-based approach. Logic models have been highlighted as useful for guiding other realist- informed evaluations and successive development of realist matrices by identifying tentative contexts, mechanisms and outcomes.11,24 By understanding what a programme intends to do, realist evaluators can assess whether this is achieved, but also why or why not.25 As NHSE did not provide a logic model in their programme specification, the present study authors aimed to devise a model to articulate programme logic and inform the realist evaluation approach.

The NHS-LCD Service Specification v1 required that service providers delivering the programme to patients “must be explicit regarding the behaviour change theory and techniques that are being used, and the expected mechanism of action of their intervention”.22 Considering the previous evidence suggesting that providers commissioned to deliver NHSE programmes are not sufficiently theory-driven,12 study authors deemed it necessary to assess whether providers’ programme designs have fidelity to this expectation.


This study aimed: (1) to develop a logic model detailing how the NHS-LCD was expected to produce behavioural changes and health outcomes, and (2) to analyse and evaluate the use and application of behaviour change theory in providers’ NHS-LCD programme designs.



Documentary analysis of the NHS-LCD Programme specification documents and the four service providers’ programme manuals, staff training materials and participant materials was conducted. The study was based on the providers of the original 10 pilot areas. The methodological approach reported by NHS-DPP evaluators was adapted;12 amendments included the absence of staff training observations and an additional extraction sheet to report the extent to which BCTs are linked to theory in providers’ design. These are described further below.

Aim 1: development of a logic model underpinning the NHS-LCD


The following programme specification documents included in the analysis were selected by authors TE and LE as they provided the basis for the NHS-LCD:


Information was extracted from the NHS-LCD service specification,22 and the DiRECT study19,22,26 detailing: the expected outcomes for patients participating in the programme; assumptions underpinning the NHS-LCD; programme structure; and behaviour change content. Information was extracted for each document separately. Further information on the BCTs specified for inclusion and the constructs targeted were extracted from NICE guidance PH6 and PH49.27,28

The information extracted from all service specification documents was firstly formulated into separate ‘If-Then’ tables for each document. Tables were then merged to form overall ‘If-Then’ statements underpinning the programme logic (Table 1). For clarity, If-Then statements were separated into two tables. The first included statements about what the service aims to achieve, whilst the second included the psychological mechanisms underpinning how the service is expected to produce change. This was followed by the formulation of a psychobehavioural logic model depicting these underlying assumptions and the psychological mechanisms linking programme structure and content to the expected short- and long-term behavioural changes and outcomes. The model was refined through team discussions and consultation with NHSE, as stakeholder input in logic model development is recommended to ensure the model accurately reflects the intervention.11,29 This process took an iterative approach, conducted over many months to refine the ‘If-Then’ table and logic model to ensure it accurately reflected the programme specification.

909 Evans Table1a 909 Evans Table1b

Aim 2: application and use of behaviour change theory in NHS-LCD providers’ programme designs

909 Evans Figure 1


The following documentation for each service provider were examined:


To identify the theoretical underpinnings of providers’ programme designs, programme content was examined using the Theory Coding Scheme (TCS).5 The TCS is a tool for assessing and reporting whether interventions are theory-based, with good inter-rater reliability evidenced by its authors.5 Documentation was sourced through liaison with the management staff of each of the four service providers. Theory coding was conducted individually and in duplicate by authors TE and CK, following author published guidance, with discrepancies resolved through discussion. Inter-rater reliability was determined using Cohen’s kappa coefficient.30 Staff training materials were sourced from the service providers. Any information on theoretical principles delivered to trainees was coded using the TCS.5

A data extraction sheet (as shown in Table 2) was developed based on the items specified in the TCS,5 including additional items proposed by NHS-DPP evaluators (items 1b, 7b, 8b and 9b). This was justified as ensuring all relevant theoretical content in programme designs are captured.12 As the service providers are not conducting the NHS-LCD evaluation, items 14-19 of the TCS were removed, as these refer to the measurement of theoretical constructs in a process evaluation - these items were also previously justified for removal by NHS-DPP evaluators.12

To assess the extent to which BCTs in providers’ programme designs were linked to theory, constructs or predictors, coders extracted information on which and how many BCTs were explicitly linked. A second data extraction sheet (Figure 1) was developed to extend upon TCS items 7a, 7b, 8a and 8b (referring to whether at least one or all BCTs have been linked to theory, a construct and/or a predictor). Once coded, data were converted into percentages to report numerically the extent to which BCTs were explicitly linked.


Logic model underpinning the NHS-LCD

The logic model describing how the NHS-LCD is expected to produce the primary outcomes of weight loss/maintenance to improve HbA1C, achieve diabetes remission and reduce medication use is shown in Figure 2.

909 Evans Figure 2

The model takes a linear format and includes two components: the theory of change and the theory of action.4 The former describes the psychological constructs targeted by the intervention that are anticipated to lead to behavioural changes, whilst the latter describes the active ingredients (BCTs) included to target these processes. Based on the information and BCTs extracted from the NHS-LCD programme specification documents, the first section of the logic model describes how information provision on the benefits of lifestyle change and social approval will lead to positive outcome expectancies (i.e. that dietary adherence will produce health benefits and reduce reliance on medication) and enhanced subjective norms (i.e. the belief that lifestyle changes are perceived positively by similar people and those to whom they aspire), resulting in programme uptake and intention to initiate behaviour change.

This section also depicts how the type of information provision changes throughout the phases of the programme: phase one focuses on the benefits of weight loss without sudden increase in exercise and on enhancing understanding of behavioural antecedents (i.e. internal and external cues of eating behaviours); phases two and three focus on behavioural instruction for both dietary and physical activity behaviours to maintain weight loss.

Section two of the model describes how participants move through a self-regulatory cycle (i.e. the ability to control their own behaviour) once phase one goals are set and TDR is initiated, including reviewing progress, receiving feedback, developing coping strategies and modifying goals in light of achievement. This facilitates self-efficacy (confidence in one’s ability to initiate and maintain behaviour change) and TDR adherence.

As participants move into phases two and three (section 3), new achievable behavioural goals are set, leading to incremental improvements in physical activity and healthy eating behaviours. Participants move into a second self-regulatory cycle that replicates the phase one cycle whilst additionally being provided with self-monitoring tools to identify and resolve discrepancies between behaviours and goals. This cycle is expected to ensure routines that support behavioural changes are developed and self-efficacy maintained.

Section four of the model describes how participants who regain 2kg or more during the weight maintenance phase return to the TDR for 4 weeks and are encouraged to problem-solve and form coping plans, ensuring that they develop self-efficacy in managing future relapses. Shaping knowledge is targeted throughout the programme through information provision, in line with UK Government guidance to support self-efficacy.31.32 A person-centred and empathy-building approach is adopted throughout, and supportive social environments are facilitated. The processes outlined were identified as mapping onto constructs proposed by three unique but overlapping models of behaviour change: the Theory of Planned Behaviour (TPB),33 the Health Action Process Approach model (HAPA)34 and Self-Regulation Theory (SRT).35

Theory described in providers’ programme plans and staff training

The theory coding analysis revealed that the application of behaviour change theory to programme designs varied across providers, summarised in Table 2. One provider explicitly linked theory to programme content by including a logic model. Two providers described the theoretical basis of their interventions in detail; however, some but not all BCTs were linked to the theories proposed. One provider had limited evidence of theory use as theory was only described on a single training slide to guide coaches’ approach to session delivery and was not linked to any programme content.

The behaviour change theories and constructs selected varied amongst providers (Table 3). Service Provider 1 included the Transtheoretical Model (Stages of Change) within their training content.36 Service Provider 2 described multiple theories as underpinning their programme design, including COM-B,37 the Health Beliefs Model,38 the Transtheoretical Model,36 and the Cognitive Behavioural Model.39,40 Service Provider 3 also included the Cognitive Behavioural model,39,40 in addition to the Antecedents-Behaviour-Consequence model.41 Finally, Service Provider 4 included COM-B and Social-Cognitive Theory in their programme logic model.37,42

909 Evans Table 2909 Evans Table 3

Analysis of the extent to which BCTs in providers’ programme designs were linked to theory, constructs or predictors revealed that Service Provider 1 linked no (0%) BCTs, Service Provider 2 linked 20 (63%) of BCTs, Service Provider 3 linked 23 (70%) of BCTs, and Service Provider 4 linked all (100%) BCTs.

Across the four providers, analysis of Cohen’s kappa revealed substantial agreement between the two researcher’s judgements when using the TCS (κ =.701, p < .001) and moderate agreement when coding linkage to BCTs (κ = .424, p < .001) prior to resolving discrepancies.43


The first aim of this study was to develop a logic model describing how the NHS-LCD programme expected to produce change. By extracting information on programme assumptions, structure and content from the NHS-LCD service specification documents, authors devised a behaviour change logic model demonstrating the active ingredients (BCTs) and the psychological mechanisms targeted throughout the three phases of the 52-week programme. Programme logic included positive outcome expectancies and subjective norms resulting from information provision, facilitating uptake and behaviour change intentions. This was followed by BCTs targeting shaping knowledge and self-regulation, alongside a supportive social environment and person-centred approach throughout the programme.

The second aim of this study was to analyse and evaluate the use and application of theory by service providers in their programme designs. Although NHSE commissioned independent service providers to design programmes that are explicit in their theory use, TCS coding revealed that although all four providers mentioned theory at least once within their designs or staff training documents, only one provider evidenced all BCTs being linked to theory and/or constructs (through construction of a logic model). Two providers linked some but not all BCTs to theory, and one provided no evidence of theory use in their programme design, aside from inclusion in a staff training slide. By additionally coding which and how many BCTs were linked to theory, constructs or a predictor, we were able to establish that the two providers evidencing some but not complete use of theory in BCT selection linked 63% and 70% of BCTs respectively, demonstrating a similar degree of theory application between them. Furthermore, six different behaviour change theories were described across all four programme designs, whilst providers appear to have utilised aspects from more than one theory to inform their programmes. Finally, no provider included details about their theory selection process, therefore if and how models were selected based on their evidence of effectiveness for dietary change and T2D management is unclear. Together this evidence demonstrates some but not complete justification for the behaviour change theories and techniques selected. The behaviour change content and fidelity of providers’ programme designs to the intervention components described in the logic model presented here will be reported in a separate paper.

Strengths and limitations

It is important to recognise that logic model construction involves an interpretative approach informed by authors’ knowledge and expertise in the health psychology domain. There may be other models or theoretical constructs that map onto the programme logic, and the logic model might have differed if constructed from outside a psychological perspective. This limitation was acknowledged and mitigated through repeated reflection throughout the logic modelling process, triangulating with co-authors’ interpretations, and an iterative process to continuously refine if-then statements and the model diagram, facilitating a robust methodology. Furthermore, as this was an independent evaluation of a national pilot programme, the research team were not involved in the programme development. Thus, the authors were in regular contact with the NHSE LCD programme management team, who confirmed this was an accurate reflection of the programme and what it set out to achieve.

All documentation describing the theoretical basis and behaviour change content of service providers’ programme designs was obtained by building successful stakeholder relationships. Providers were given the opportunity over many months to provide all documentation, including any existing logic model. Our methodological approach was informed by that set out by NHS-DPP evaluators,11 including the use of a validated tool for coding theory use,5 ensuring clear programme comparisons can be made by researchers and stakeholders. However, it was not possible to analyse staff training in a way comparable to the NHS-DPP evaluation, as staff training courses had already taken place prior to the commissioning of this evaluation and thus could not be observed. Therefore, the extent to which staff were trained in the theoretical principles underpinning their programme could not be fully assessed.

Relation to existing research

Evaluators of the NHS-DPP found a lack of explicit theory use in the behaviour change content across providers delivering the programme.12 Our findings are similar in that most providers did not evidence explicit use and application of theory. One provider did include a logic model to describe how their intervention expected to achieve the desired programme outcomes; however, no justification was documented for the theories and constructs selected for the purposes of the NHS-LCD programme. This is an important omission as the appropriateness of a behaviour change model for any given target behaviour or population cannot be assumed; selection must be based on evidence demonstrating effectiveness for improving dietary behaviours and T2D management. These findings demonstrate that improvements are still needed for ensuring that large-scale behaviour change programmes are clear in their theoretical underpinnings and what constructs they are targeting, and why.

Additional similarities exist between the NHS-DPP12 and NHS-LCD logic models, with both having elements that map onto the HAPA model34 and include a self-regulatory cycle.35 This is not surprising considering they are both national NHS programmes that have taken a similar approach in their commissioning model, and are based on similar behaviour change evidence for similar behaviours and populations. However, in contrast to the NHS-DPP model,12 initial information provision in the NHS-LCD targets positive outcome expectancies (for example, the belief that behaviour change will reduce reliance on medications) as opposed to risk perception (for example, the belief that one is at risk of future reliance on medication if behaviours remain unchanged). This is in line with evidence suggesting that targeting intrinsic motivation (the person’s own desire or intention to perform a behaviour) is related to better diabetes self-care, whilst fear appeals may lead to avoidance or denial for people living with diabetes.44 Furthermore, unlike the NHS-DPP, the NHS-LCD implements multiple programme phases, across which the target behaviours and outcomes change to facilitate maintenance of treatment outcomes.

909 Evans Key Messages

Implications and recommendations

The findings of this study add to the growing body of research demonstrating that theory is not always explicitly and effectively applied throughout the entirety of the intervention development process, including the design of diabetes programmes specifically.5,7-10,12 This is in line with findings of the NHS-DPP evaluation, in that they support the notion that health promotion interventions are often ‘evidence-inspired’ rather than a rigorous scientific process.5,12 Michie and Prestwich argued that although theory is included in numerous frameworks, a detailed guide to its use during the design process is lacking and would benefit those without expertise in theory use.5 The MRC has addressed this in its recently updated framework for designing and evaluating complex interventions,1 which guides developers through the iterative process of applying programme theory throughout all development phases.

More needs to be done to achieve the rigorous application of theory in national programmes: future research and consultancy work should seek to understand and address the barriers to explicit theory use in programme design. For example, it is important to consider the time and resource constraints that are likely experienced by intervention developers, especially within the context of health services, which may act as a barrier to a meticulous theoretically driven process. Providers often have previously established intervention components found to be effective in other programmes; providers may adapt these components instead of using a theoretically informed approach, thus reducing staff training and resource development needs. However, it is best practice to ensure that theory guides intervention development, identification of target constructs and selection of appropriate BCTs to target these mechanisms, and not to source theory to justify pre-designed intervention components.1

Similar to conclusions reached by NHS-DPP evaluators,12 NHSE guidance on theory use in the programme specification was vague, with only the instruction to be explicit in theory application and to use the Public Health England Behaviour Change Guide.45 No specific guidance on the target constructs of the programme was provided, and although study authors were able to unpick these through a meticulous logic model construction process, this would be time-consuming for service providers to achieve. The lack of guidance may explain the variation in behaviour change theories and constructs selected by providers, and will likely result in variation of behaviour change content across their designs – a finding observed in the NHS-DPP.12 However, it was the intention of NHSE to ensure providers had scope to design their own programmes based on their own behaviour change expertise. It is therefore recommended that outcomes across providers are compared to evaluate whether differences in the theoretical approaches adopted are associated with better outcomes.

Without a clear underpinning theory, providers do not have a clear justification of the intervention techniques they plan to use in their programme designs. This variation and lack of clear under-pinnings may result in a loss of fidelity in behaviour change content and its delivery,12 and may result in variation in what is received by programme recipients and thus variation in programme outcomes. This may be problematic considering that interventions requiring high personal agency, such as weight management programmes, are argued potentially to exacerbate existing health inequalities. Inclusion of the logic model constructed in the NHSE specification may improve theory use amongst providers and fidelity to the programme specification. Alternatively, future NHSE commissioning processes should require providers to demonstrate a) explicit theory use (i.e., by providing a logic model), and b) the process in which appropriate theory for the target group and health behaviours were selected. Alongside this, more detailed guidance or training in theory application should be included, to support them in achieving this.

Finally, programme theory, explicitly communicating how an intervention expects to produce change, is essential for conducting a thorough realist evaluation. By developing a logic model, evaluators of the NHS-LCD programme can identify the proposed mechanisms underpinning why or why not the programme works and assess the impact of contexts on these mechanisms.4


The NHS-LCD aims to improve T2D self-management and reliance on medication use by providing behavioural support that facilitates TDR adherence, healthy lifestyle behaviours and resulting weight loss/maintenance. Despite NHSE commissioning providers to design behaviour change interventions explicit in their theoretical underpinnings and mechanisms of action, only one of the four providers evidenced a detailed description of how their intervention was expected to achieve the desired outcomes through the construction of a logic model. Without a clear underpinning programme theory describing how providers programmes expect to produce behavioural changes and health outcomes, justification for the BCTs selected is unclear and may result in a drift in the fidelity of programme delivery.12 To prevent this drift, it is recommended that programme commissioners adopt the logic model presented in this manuscript or require service providers to state their underpinning programme theory explicitly during the commissioning process to ensure a rigorous, theoretically-driven approach to developing and implementing national programmes.

Conflict of interest The authors declare that they have no competing interests.

Funding The National Institute for Health Research [NIHR132075]. The funding body did not take any role in the design of the study, data collection and analysis, or in writing the manuscript.

>Availability of data and materials The datasets analysed during the current study are available from the corresponding author on reasonable request.

Authors’ contributions TE lead the conceptualisation, design, analysis, interpretation of data, and drafted the report for this study. RH made substantial contributions to the interpretation of data and revision of the report. CK made substantial contributions to the data analysis for this study. LN made substantial contributions to the interpretation of data and revision of the report. DR made substantial contributions to data collection and revision of the report. AH and JM made substantial contributions to the revision of the report. LE made substantial contributions to the design of this study and revision of the report and supervised TE. All authors read and approved the final manuscript.

Acknowledgments The Re:Mission research team and the NHS-LCD Programme management team.


  1. Skivington K, Matthews L, Simpson SA et al. A new framework for developing and evaluating complex interventions: update of Medical Research Council guidance. BMJ 2021;374:n2061.
  2. Rimmer BK, Glanz K. Theory at a glance: A guide for health promotion . Washington. DC: National Institutes for Cancer, Department of Health and Human Services. 2005.
  3. Glanz K. Theory at a glance: A guide for health promotion practice: US Department of Health and Human Services, Public Health Service, National Institutes of Health, National Cancer Institute; 1997.
  4. Funnel SC, Rogers PJ. Purposeful program theory: Effective use of theories of change and logic models. John Wiley & Sons; 2011;1:95-148.
  5. Michie S, Prestwich A. Are interventions theory-based? Development of a theory coding scheme. Health Psychology 2010;29(1):1.
  6. Escoffery C, Lebow-Skelley E, Haardoerfer R et al. A systematic review of adaptations of evidence-based public health interventions globally. Implementation. Science 2018;13(1):1-21.
  7. Glanz K, Bishop DB. The role of behavioral science theory in development and implementation of public health interventions. Ann Review Public Health 2010;31:399-418.
  8. McSharry J, Byrne M, Casey B et al. Behaviour change in diabetes: behavioural science advancements to support the use of theory. Diabetic Medicine 2020; 37(3):455-63.
  9. Painter JE, Borba CPC, Hynes M et al. The use of theory in health behavior research from 2000 to 2005: a systematic review. Annals Behavioral Medicine 2008;35(3):358-62.
  10. Prestwich A, Sniehotta FF, Whittington C et al. Does theory influence the effectiveness of health behavior interventions? Meta-analysis. Health Psychology 2014;33(5):46.
  11. Ebenso B, Manzano A, Uzochukwu B et al. Dealing with context in logic model development: reflections from a realist evaluation of a community health worker programme in Nigeria. Evaluation Program Planning 2019; 73:97-110.
  12. Hawkes RE, Miles LM, French DP. The theoretical basis of a nationally implemented type 2 diabetes prevention programme: how is the programme expected to produce changes in behaviour? Int J Behavioral Nutrition Physical Activity 2021;18(1):1-12.
  13. Greenwell K, Sivyer K, Vedhara K et al. Intervention planning for the REDUCE maintenance intervention: a digital intervention to reduce reulceration risk among patients with a history of diabetic foot ulcers. BMJ Open 2018; 8(5):e019865.
  14. Lake AJ, Browne JL, Abraham C et al. A tailored intervention to promote uptake of retinal screening among young adults with type 2 diabetes-an intervention mapping approach. BMC Health Services Research 2018; 18(1):1-18.
  15. NHS England. Low calorie diets to treat obesity and Type 2 diabetes. 2020. Accessed 12 Feb 2022.
  16. Steven S, Lim EL, Taylor R. Population response to information on reversibility of type 2 diabetes. Diabetic Medicine 2013;30(4):e135-e8.
  17. Steven S, Hollingsworth KG, Al-Mrabeh A et al. Very low-calorie diet and 6 months of weight stability in type 2 diabetes: pathophysiological changes in responders and nonresponders. Diabetes Care 2016;39(5):808-15.
  18. Lim EL, Hollingsworth KG, Aribisala BS et al. Reversal of type 2 diabetes: normalisation of beta cell function in association with decreased pancreas and liver triacylglycerol. Diabetologia 2011;54(10):2506-14.
  19. Lean MEJ, Leslie WS, Barnes AC et al. Primary care-led weight management for remission of type 2 diabetes (DiRECT): an open-label, cluster-randomised trial. Lancet 2018;391(10120):541-51.
  20. Lean MEJ, Leslie WS, Barnes AC et al. Durability of a primary care-led weight-management intervention for remission of type 2 diabetes: 2-year results of the DiRECT open-label, cluster-randomised trial. Lancet Diabetes Endocrinol 2019;7(5):344-55.
  21. Churuangsuk C, Hall J, Reynolds A et al. Diets for weight management in adults with type 2 diabetes: an umbrella review of published meta-analyses and systematic review of trials of diets for diabetes remission. Diabetologia 2022;65:14-36.
  22. NHS England. Service Specification No. 1: NHS Low Calorie Diet Programme. [Version 01]. 2021.
  23. Hawkes RE, Cameron E, Bower P, French DP. Does the design of the NHS diabetes prevention programme intervention have fidelity to the programme specification? A document analysis. Diabetic Medicine 2020;37(8):1357-66.
  24. Renger R, Foltysova J, Becker KL, Souvannasacd E. The power of the context map: Designing realistic outcome evaluation strategies and other unanticipated benefits. Evaluation Program Planning 2015;52:118-25.
  25. Rogers PJ. Using programme theory to evaluate complicated and complex aspects of interventions. Evaluation 2008;14(1):29-48.
  26. Leslie WS, Ford I, Sattar N et al. The Diabetes Remission Clinical Trial (DiRECT): protocol for a cluster randomised trial. BMC Family Pract 2016;17(1):1-10.
  27. National Institute for Health and Care Excellence. Behaviour change: general approaches [Internet]. [London]: NICE, 2007 [updated 2017 Dec]. (Public Health Guideline [PH6]). Available from: Guidance/PH6.
  28. National Institute for Health and Care Excellence. Behaviour change: individual approaches [Internet]. [London]: NICE, 2014. (Public Health Guideline [PH49]). Available from:
  29. Funnell SC, Rogers PJ. Purposeful program theory: Effective use of theories of change and logic models. John Wiley & Sons; 2011.
  30. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159-74.
  31. Public Health England. The Eatwell Guide. [Internet]. 2016 [updated 2018]. Available from:
  32. Department of Health and Social Care. Physical activity guidelines: UK Chief Medical Officers' report. [Internet. 2019 [updated Sep 2020]. Available from:
  33. Ajzen, I. (1985) From intentions to action: a theory of planned behavior. In: Kuhl J and Beckmann J, Eds, Action-Control: from cognition to behavior, Springer-Verlag, Heidelberg, 11-39.
  34. Schwarzer R. Modeling health behavior change: How to predict and modify the adoption and maintenance of health behaviors. Applied Psychology 2008;57(1):1-29.
  35. Cs C. Scheier MF. On the self-regulation of behaviour. Cambridge, UK: Cambridge University Press; 1998.
  36. Prochaska JO, Velicer WF. The transtheoretical model of health behavior change. Am Journal Health Promotion 1997;12(1):38-48.
  37. Michie S, Atkins L, West R. The behaviour change wheel: a guide to designing interventions. Silverback Publishing; 2014.
  38. Rosenstock IM. Why people use health services. Milbank Quarterly 2005;83(4).
  39. Beck A. Cognitive therapy and the behavioral disorders. New York: Meridian; 1967.
  40. Beck AT. Cognitive therapy: Nature and relation to behavior therapy. J Psychotherapy Practice Research 1993;2(4):342.
  41. Bijou SW, Peterson RF, Ault MH. A method to integrate descriptive and experimental field studies at the level of data and empirical concepts 1. J Applied Behavior Analysis 1968;1(2):175-91.
  42. Bandura A. Social foundations of thought and action. Englewoods Cliffs. NJ: Prentice-Hall; 1986.
  43. McHugh ML. Interrater reliability: the kappa statistic. Biochemia medica 2012;22(3):276-82.
  44. Swanson V, Maltinsky W. Motivational and behaviour change approaches for improving diabetes management. Practical Diabetes 2019;36(4):121-5.
  45. Public Health England. Behaviour change: guides for national and local government and partners. [Internet]. 2020. Available from: government/publications/behaviour-change-guide-for-local-government-and-partners.