What was successful and not successful in the programme
The research programme has been delivered largely as planned, with all trials recruiting to target and with good retention and follow-up. We generated robust evidence of the cost-effectiveness of scalable BWMPs with clear recommendations for policy-makers about the value of these interventions for people at high risk of T2D and those with a recent diagnosis. These have been disseminated directly to policy-makers and practitioners at a national and regional level. We have published 24 scientific articles and key findings have been published in leading journals in the field (e.g. Lancet Public Health, Diabetologia, Obesity Reviews, International Journal of Obesity). We have also received several awards based on this research (RCGP Paper of the Year in Health Services Research for WRAP 5-year follow-up; Association for the Study of Obesity (ASO) Prevention and Treatment of Obesity Award and Foster Schauer Travel Grant for SWiM COVID; Diabetes UK Travel Award for GLoW trial). Following the successful feasibility study for SWiM, we also secured a NIHR Programme Grant for Applied Research to further develop and evaluate SWiM.
We did face a number of unexpected challenges, most notably the COVID-19 pandemic which came at a critical time for the studies within the programme, specifically the GLoW trial (WP1.2) and the SWiM feasibility study (WP2.2).
The Glucose Lowering through Weight management trial
For GLoW, we planned to recruit 576 participants over 10 months using 30 GP sites. We quickly recruited 30 practices, but initial feasibility searches had overestimated the number of eligible patients per practice and mailouts yielded far fewer enrolments than anticipated. There were often delays in recruited practices sending invites, meaning patients had already been referred to diabetes education and were no longer eligible. We worked closely with the Clinical Research Network (CRN) to identify new GP sites and used monthly newsletters and incentives for practice staff to encourage timely mailouts and increase opportunistic recruitment. We introduced new processes to reduce practice burden (e.g. an automated mailing service). When the first COVID lockdown occurred in March 2020 we had recruited 159 GP sites and were on track to complete GLoW within the programme timeline and without increasing the budget.
The GLoW trial was then paused by the sponsors between March 2020 and January 2021 because CRNs and primary care practices were unable to support non-COVID-related research activity. We worked with our Cambridge Epidemiology and Trials Unit and the CRN to adapt study protocols and develop remote data collection methods that maintained high standards of data quality and minimised data loss during the pause. This included systems for collecting self-reported outcomes via electronic questionnaires, and remote collection of finger prick blood samples for measuring HbA1c. GP practices were asked to complete notes reviews in place of the usual clinic visits. An e-consent system was introduced to enable remote recruitment for those reluctant to attend clinic when we were allowed to restart in January 2021.We completed recruitment to target on 6 August 2021. However, this meant that timelines for data collection shifted considerably and we had to request a costed extension to enable delivery of the trial.
In addition to recruitment and retention issues, the pandemic interrupted delivery of the intervention and standard care. In-person attendance to WW and DESMOND classes ceased, moving to web-based video calls (WW) and the MyDESMOND App (https://www.mydesmond.com/home/) instead. Data on intervention delivery method was captured via self-report and records from WW and DESMOND providers. An in-depth process evaluation has enabled us to understand and account for the disruption to intervention delivery during that period.
Supporting Weight Management feasibility study
At the time of lockdown we had prepared the ethics application for the SWiM feasibility study but were advised by the sponsors that this should not be submitted to Health Research Authority for approval as it was not a COVID-related research project and could not be supported at that time. To make use of the period of delay, we undertook further rounds of user testing. We also responded to an urgent need for weight management support during the COVID-19 lockdown that was communicated to us by organisations representing people living with obesity (European Patient Council on Obesity, Obesity UK). We adapted the SWiM intervention to support people with overweight and obesity during the COVID pandemic and with additional funding from the European Association for the Study of Obesity, we ran a remote trial of this new intervention (SWiM-C). Findings were published in Mueller et al.79,80 and we were awarded the ASO Prevention and Treatment of Obesity Award for this research. Findings provided additional insight into the feasibility of the SWiM intervention, which were used to inform intervention refinement for when the feasibility study commenced.
Patient and public involvement
Patient and public involvement was integrated throughout the research programme. Our PPI lay coinvestigator helped lead the PPI input and attended and presented at investigator meetings. PPI was included as a standing agenda item on all project and programme meetings. We encouraged all team members to be proactive and reactive in identifying where PPI adds value their work.
Patient and public involvement representatives
We recruited a diverse patient/public user group panel (PUGP) who had lived experience of obesity or T2D. The panel consisted of 16 members, of whom 9 regularly responded and attended meetings. Members of the PUGP were recruited through local diabetes support groups and weight management groups, as well as previous participants in the WRAP study.
Patient and public involvement capacity building
In the spirit of reciprocal learning we involved junior and senior research staff in our PPI in order to build their skills and enthusiasm for PPI. We used our PUGP meetings to enable junior staff to have a safe and supportive environment to present WPs and findings and have shared PPI good practice through the levels of the research team to leave a lasting legacy for PPI in future projects. In terms of PUGP members, we have encouraged involvement in other allied projects and have been able to recommend and appoint someone from the PUGP as a co-applicant on another successful grant. In adopting this reciprocal approach to PPI we are building capacity for involvement in public health research in the future both within our unit and in the wider research environment.
Intervention development
A key focus of the PUGP was involvement in the design of the weight loss maintenance intervention to ensure acceptability, engagement and feasibility for users (WP2.1). Regular coproduction workshops enabled them to input into every step of the planning process: the logic model, the outline, different prototype iterations. Feedback from the PUGP helped to shape the intervention in terms of content, delivery and functionality. A key example of the changes we made included not giving specific dietary advice. Target users have already lost weight and have strong opinions on what works for them. We decided to build a programme that builds on that rather than working against it. We also added a specific module supporting healthy sleep and a weight tracker, which we had been unsure about including previously. The font size throughout the web platform was increased and pages restructured to have fewer words per page, and content was adapted to focus on actions and activities rather than theory. The module on ‘Values’ was completely revised through dialogue with the PUGP to make it more readily understandable. To get feedback on the content from a wider and more diverse audience, we conducted remote user testing of the web platform with the PUGP members and participants from the WRAP study (WP1.1) from around the UK.
Input to trials
For all the trials within the programme, PPI input was sought to maximise participant engagement and retention, and minimise measurement burden. The PUGP were given the opportunity to comment on trial design and review the content, design, and delivery of all participant-facing materials. PPI representatives also advised on interview schedules for the qualitative interviews and focus groups to ensure sensitivity, and maximise participant engagement. The PPI co-applicant coauthored trial protocols.
Governance
The Programme Steering Committee included two PPI representatives who actively participated in meetings, reviewed documents and provided input to the programme in-between meetings. Our research ethics submission was aided by letters of support from our PUGP members and also by identifying ethical issues and suggesting mitigation strategies from a participant perspective.
Results and dissemination
The findings of all studies were presented to the PUGP to aid interpretation of findings. In WP1.1, 1.2 and 2.2 a PPI representative took part in the analysis, interpretation and write-up of findings for the qualitative study, coauthoring the resulting publications. The PPI co-applicant also coauthored a number of publications arising from the trial, including the primary trial papers.
Patient and public involvement input was sought in the development of dissemination plans, ensuring an appropriate range of audiences were targeted and messages were communicated effectively. PPI representatives advised on content and methods of dissemination and reviewed public facing documents such as newsletters and press releases. Working closely with patient organisations (ECPO, Obesity UK) has also helped to widen the reach of our dissemination activities.
Strengths and challenges of the patient and public involvement and engagement
One of the key strengths of our PPIE work was the strong relationships we built with a diverse group of people with lived experience. Having a strong PPI chair for PUGP meetings with the Investigator team supported PPI contributors to feel empowered to give critical feedback. We also drew on other strengths of our PPIE panel. For example, members with a background in education gave important advice on how to communicate complex information, and members with information technology backgrounds were able to communicate directly with our software developers.
The main challenge we faced was not being able to run face-to-face meetings during the COVID-19 pandemic. We observed that some of our original group members from underserved groups did not attend meetings after we moved to video conferencing. We maintained regular communication with these members via e-mail and telephone, and offered telephone calls as an alternative to meetings to ensure that their voice was still heard. Despite some members not attending meetings, because of the large group recruited we have been able to ensure that diverse voices are represented at all meetings. We are fortunate to have had such a loyal and supportive core membership despite the challenges that the pandemic has brought. Moving forward, we aim to conduct hybrid meetings and anticipate that this will improve attendance of original group members at PUGP meetings. It is also worth noting that a research programme (as opposed to a distinct project) has, by virtue of the length of the programme, challenges with maintaining an active PPI group. We have managed to insulate ourselves from this common issue with the support of our PPI lead and the enthusiasm and engagement of the principal investigator and research team.
Training and support
We keep an open dialogue with our PPI advisors to review training and support needs. We have a reciprocal approach to training whereby the research team learn from the lay advisors and vice versa; in doing so, we enable a holistic environment which we hope will strengthen the skills and expertise of all working on this programme. One of the group members is now a co-applicant on another project and is being mentored by Jennifer Bostock which we see as a core product of our PPI work to build a pool of lay advisors to engage in future work and to act as ambassadors for our outputs, our methods and our research programmes.
The key challenges of this work were trying to integrate diverse and often conflicting opinions and balance these with other important factors, such as the scientific evidence about what works, constraints around logistics, ethics, and data security, as well as our budget (both in terms of the development and how much it would cost to roll out). This required a lot of discussion and negotiation. By having some focus groups with the research team, the PPI group and the software developers all together, we could identify alternative ways to implement some suggestions. At each meeting, we fed back about which changes we had made since the last meeting and explained why some changes had not been implemented. It was also difficult to keep a diverse group engaged over the long term.
Equality, diversity and inclusion
A strength of this research programme is the use of minimal exclusion criteria in each of the trials, with all eligible participants invited to take part. To increase inclusion, all protocols and participant-facing materials were created with a diverse group of people with lived experience. Samples for WRAP and GLoW were broadly generalisable to the target population, with good diversity in gender and socioeconomic status. However, ethnic diversity reflects the makeup of the UK population (≈90% White), which limits generalisability to other ethnic groups. In the SWiM feasibility study, participants were mostly female and White, which did not reflect the target population. Some of these inequalities may be inherited from the programmes we were recruiting from. Future research evaluating SWiM must focus on recruiting a more diverse sample.
We collected detailed demographic data for all participants, guided by PROGRESS Plus criteria. We have analysed intervention effectiveness and cost-effectiveness by key demographic characteristics and we will make data available for secondary analyses and individual participant data meta-analyses. We have separately undertaken our own individual participant meta-analysis of the moderation of the behavioural weight loss programmes by demographic characteristics.53
The research team represented good gender, age and socioeconomic diversity and a range of experience and expertise. We had a strong focus on supporting career development. Most outputs were first-authored by early-mid career researchers and postgraduate students, and several early career researchers on this programme grant are WP leads and/or co-applicants on our new programme grant.
Impact and learning
This research programme has generated robust evidence on the long-term effectiveness and cost-effectiveness of scalable BWMPs, directly informing national and regional policy discussions on diabetes prevention and management. Beyond the directly NIHR PGfAR-funded work, this programme has contributed data to seven PhD theses and five MPhil theses on related research (University of Cambridge, University of Oxford, University of Liverpool, University of Sheffield and University of Glasgow). It has also contributed data to four individual participant data meta-analyses: two led by our research group, one led by the University of Oxford, one led by University of Lancaster. Secondary data use is promoted and data dictionaries, including detailed meta-data, are publicly available.
The research findings have been disseminated to policy-makers, professional societies, healthcare providers and the research community and longer term this work is expected to enhance the effectiveness, efficiency and equity of behavioural weight management provision across the NHS.
This programme contributes to sustainable health systems by focusing on scalable, cost-effective behavioural interventions that aim to reduce the long-term burden of obesity and T2D, thereby supporting NHS efficiency and environmental sustainability through reduced resource use in chronic disease management. Remote delivery of our interventions and the move towards remote data collection in our trials (accelerated during the COVID-19 pandemic) reduced travel and carbon emissions during the trial and demonstrated the feasibility of this sustainable approach. The planned cost-effectiveness trial of SWiM will use entirely remote methods.
Implications for decision-makers
Findings from this research programme indicate that scalable BWMPs could represent a cost-effective use of NHS and public sector resources for people at high risk of, or recently diagnosed with, T2D. Evidence from WP1.1 to WP1.3 supports the inclusion of such programmes within diabetes prevention and management pathways, provided stated limitations relating to generalisability, and future research needs are considered. The development and feasibility testing of the SWiM intervention further provide a model for low-cost, scalable approaches to support weight loss maintenance following behavioural programmes, and the clinical and cost-effectiveness of this approach is currently being evaluated in collaboration with national and regional decision-makers.
Research recommendations
- Further research is warranted to design, optimise and evaluate interventions that promote long-term weight loss maintenance. Such work should specifically address the psychological and behavioural processes that underpin sustained weight management.
- The SWiM intervention shows promise for supporting long-term outcomes in people who have lost weight through a behavioural programme. Further research is needed to refine the SWiM intervention and undertake a definitive trial to evaluate its long-term effectiveness and cost-effectiveness as a scalable solution for supporting weight loss maintenance. This work is currently ongoing and will be crucial for determining its potential for wider implementation within NHS and public health settings.
- Future investigations into scalable behavioural weight management approaches should focus on improving accessibility, inclusivity and participant engagement. In particular, research should explore strategies to overcome digital exclusion, enhance cultural tailoring and perceived relevance, and strengthen trust and credibility among groups who are currently under-represented or disadvantaged.
- Ongoing work should prioritise the refinement and validation of long-term health economic models that incorporate empirically observed weight change trajectories. Integrating real-world data will enable more accurate estimation of lifetime cost-effectiveness. Research should facilitate the evaluation of multiple intervention strategies to support weight loss and maintenance and consider how benefits and costs are distributed across different socioeconomic and demographic groups.
- More researchers should make use of health economic modelling as a part of the design of interventions and their evaluations to improve the likelihood that they will offer value for money to the healthcare system.
Conclusions
This research programme provides high-quality evidence from RCTs to support the use of scalable BWMPs as a cost-effective intervention for the prevention and early treatment of T2D. While absolute weight losses are more modest than those achieved by more intensive and expensive options (e.g. total diet replacement, medication, surgery), health economic modelling demonstrates that scalable behavioural programmes are very cost-effective (potentially cost saving from a lifetime perspective) and could have a substantial public health impact.
Our long-term follow-up of the WRAP trial suggests that weight regain following a commercially available BWMP has often been overestimated. However, weight regain after weight loss is still substantial and has significant impacts on cost-effectiveness. Our findings support greater investment in behavioural interventions of longer treatment duration than the 12 weeks typically offered by the NHS and local authorities, and development and commissioning of specific support for weight loss maintenance.
Our findings highlighted that many people who regain weight following a behavioural programme could benefit from additional support to address the psychological challenges of maintaining weight. We identified some cognitive and behavioural strategies that support better long-term weight management and that should be included in interventions to support weight loss maintenance and we developed a novel weight loss maintenance intervention. This intervention builds on an existing evidence base that incorporating strategies from ACT can improve long-term weight outcomes and support improvements in other psychosocial outcomes.
The SWiM intervention we have developed demonstrates considerable promise and our findings support the further development and evaluation of SWiM. We have recently obtained NIHR PGfAR funding to support the further development and evaluation of SWiM.