Research & publications
Across more than two decades, this body of work spans the full arc of behaviour change science, from developing and evaluating the MEND programme to building and testing large digital behaviour change platforms in obesity management. Across RCTs, population studies, and real-world deployments, the common thread is understanding what it takes to change behaviour at scale and doing it safely.
View full ORCID record27
Peer-reviewed works on ORCID
24
Years of published research, 2002 to 2026
16
Academic institutions collaborated with, across six countries
Featured research
The MyBarriers Questionnaire (MBQ) study was designed to identify cognitive behavioural phenotypes among adults initiating GLP-1 receptor agonist-based therapy within a digital weight management programme.
The Sacher AI team led the research in collaboration with the internal Numan team across two sequential studies, with validation in more than 400 participants, producing a structured, psychometrically evaluated, validated self-report tool.
Presented at the European Congress of Obesity (ECO) 2026. Under peer review.
Research with partners
Behavioural phenotyping in GLP-1 care
The Sacher AI team led the phenotype study behind Numan's GLP-1 weight management programme, in collaboration with Numan, and includes named inventors on a pending US patent application for the resulting digital phenotyping system, which identifies the behavioural barriers people face when losing weight.
Digital behaviour change and remote coaching
Led the behavioural science and research behind Allurion's text-based digital coaching, published as a peer-reviewed paper in JMIR Formative Research. In a 12-month study of 107 patients on a swallowable gastric balloon programme, the coaching was associated with significant improvements in mental wellbeing (P=.002) and weight-related quality of life (P<.001); 82% of patients found it helpful and 91% felt confident about making changes. Related work applied machine learning to flag likely non-responders early, so clinicians could intervene when it mattered most.
Read the peer-reviewed paper (JMIR Formative Research) ↗Family-based obesity in childhood, evidenced at scale
Co-founded MEND and led its research programme, including randomised controlled trials of a family-based intervention for obesity in childhood, delivered across communities in the UK and the US.
MEND randomised controlled trial, Obesity 2010 ↗Behaviour change at population scale
Implementation and translation lead on NoHoW, a European Horizon 2020 randomised controlled trial of digital tools for weight-loss maintenance. Led the Slimming World work package and was a key member of the study team, working with a cohort of more than 2,000 successful maintainers at one of Europe's largest community weight-management services.
The NoHoW project, Obesity Facts 2021 ↗Research collaboration
Paul Sacher is co-founder and Research Director of the Behavioral AI Institute, an independent nonprofit advancing behavioural science as a core discipline in the design and evaluation of AI. With the Institute we co-authored the open letter The missing discipline in AI: a call for behavioural science in Wellcome Open Research, alongside leading behavioural scientists and now backed by more than 500 signatories, with further research in preprint.
Read the open letter in Wellcome Open Research ↗AI safety and oversight
In a multi-country evaluation of a patient-facing AI agent, spanning more than 1,300 conversations across 30 clinics in 9 countries and scored with our published FAST framework, around one in fifteen patient messages had to be escalated to a human clinician for a potential clinical or weight-loss concern. Patient-facing AI will always surface moments that need a person. That is exactly why we build safety evaluation and human oversight in from the start, rather than bolting them on later.
Publications
2018-2026
Wellcome Open Research 2026
Sacher PM et al.
Frontiers in Digital Health 2025
Neary M, Fulton E, Rogers V, Wilson J, Griffiths Z, Chuttani R, Sacher PM
JMIR Formative Research 2024
Sacher PM, Fulton E, Rogers V, Wilson J, Gramatica M, Dent JE, Aarts EO, Eccleston D, Greve JW, Palm-Meinders I, Chuttani R
Health Education and Behavior 2018
Hingle M, Patrick H, Sacher PM, Castro Sweet C
2024
European Congress on Obesity (ECO) 2024
Sacher PM, Wilson J, Rogers V, Neary M, Fulton E
Conference abstract
European Congress on Obesity (ECO) 2024
Sacher PM, Neary M, Rogers V, Wilson J, Fulton E
Conference abstract
ASMBS Annual Meeting 2024
Sacher PM, Neary M, Rogers V, Wilson J, Fulton E
Conference abstract
2023
Obesity 2023
Ainscough BJ, Sacher PM, Martinez C, Feldgoise J
Conference abstract · doi.org/10.1002/oby.23939
Obesity 2023
Sacher PM et al.
Conference abstract · doi.org/10.1002/oby.23939
Obesity 2023
Rogers V, Sacher PM et al.
Conference abstract · doi.org/10.1002/oby.23939
Canadian Obesity Summit 2023
Sacher PM, Fulton E, Rogers V, Wilson J, Gramatica M, Liu J, Chuttani R
Conference abstract
Canadian Obesity Summit 2023
Fulton E, Sacher PM, Griffiths S, Rogers V, Wilson J, Chuttani R
Conference abstract
2019-2021
Obesity Facts 2021
Stubbs R, Duarte C, Palmeira A, Sniehotta F, Horgan G, Larsen S, Marques M, Evans E, Ermes M, Harjumaa M, Turicchi J, O'Driscoll R, Scott S, Pearson B, Ramsey L, Mattila E, Matos M, Sacher P, Woodward E et al.
Digital Health 2020
Messiah SE, Sacher PM, Yudkin J, Ofori A, Qureshi FG, Schneider B, Hoelscher DM, de la Cruz-Muñoz N, Barlow SE
Obesity 2019
Messiah SE, Sacher PM, Yudkin J, Qureshi FG, Hoelscher DM, Barlow SE
2010-2018
International Journal of Obesity 2018
Sacher PM, Kolotourou M, Poupakis S, Chadwick P, Radley D, Fagg J
Obesity (Silver Spring) 2017
Butte NF, Hoelscher DM, Barlow SE, Pont S, Durand C, Vandewater EA, Liu Y, Adolph AL, Perez A, Wilson TA, Gonzalez A, Puyau MR, Sharma SV, Byrd-Williams C, Oluyomi A, Huang T, Finkelstein EA, Sacher PM, Kelder SH
Childhood Obesity 2015
Kolotourou M, Radley D, Gammon C, Smith L, Chadwick P, Sacher PM
Obesity 2010
Sacher PM, Kolotourou M, Chadwick PM, Cole TJ, Lawson MS, Lucas A, Singhal A
Full publication record on ORCID (0000-0002-4374-7310).
Next step
If you need research, evaluation, or stronger external credibility for your AI, book a discovery call.
