Sandforth A, von Schwartzenberg RJ, Arreola EV, Hanson RL, Sancar G, Katzenstein S, Lange K, Preißl H, Dreher SI, Weigert C, Wagner R, Kantartzis K, Machann J, Schick F, Lehmann R, Peter A, Katsouli N, Ntziachristos V, Dannecker C, Fritsche L, Perakakis N, Heni M, Nawroth PP, Kopf S, Pfeiffer AFH, Kabisch S, Stumvoll M, Schwarz PEH, Hauner H, Lechner A, Seissler J, Yurchenko I, Icks A, Solimena M, Häring HU, Szendroedi J, Schürmann A, de Angelis MH, Blüher M, Roden M, Bornstein SR, Stefan N, Fritsche A, and Birkenfeld AL
Background: Remission of type 2 diabetes can occur as a result of weight loss and is characterised by liver fat and pancreas fat reduction and recovered insulin secretion. In this analysis, we aimed to investigate the mechanisms of weight loss- induced remission in people with prediabetes., Methods: In this prespecified post-hoc analysis, weight loss-induced resolution of prediabetes in the randomised, controlled, multicentre Prediabetes Lifestyle Intervention Study (PLIS) was assessed, and the results were validated against participants from the Diabetes Prevention Program (DPP) study. For PLIS, between March 1, 2012, and Aug 31, 2016, participants were recruited from eight clinical study centres (including seven university hospitals) in Germany and randomly assigned to receive either a control intervention, a standard lifestyle intervention (ie, DPP-based intervention), or an intensified lifestyle intervention for 12 months. For DPP, participants were recruited from 23 clinical study centres in the USA between July 31, 1996, and May 18, 1999, and randomly assigned to receive either a standard lifestyle intervention, metformin, or placebo. In both PLIS and DPP, only participants who were randomly assigned to receive lifestyle intervention or placebo and who lost at least 5% of their bodyweight were included in this analysis. Responders were defined as people who returned to normal fasting plasma glucose (FPG; <5·6 mmol/L), normal glucose tolerance (<7·8 mmol/L), and HbA 1c less than 39 mmol/mol after 12 months of lifestyle intervention or placebo or control intervention. Non-responders were defined as people who had FPG, 2 h glucose, or HbA 1c more than these thresholds. The main outcomes for this analysis were insulin sensitivity, insulin secretion, visceral adipose tissue (VAT), and intrahepatic lipid content (IHL) and were evaluated via linear mixed models., Findings: Of 1160 participants recruited to PLIS, 298 (25·7%) had weight loss of 5% or more of their bodyweight at baseline. 128 (43%) of 298 participants were responders and 170 (57%) were non-responders. Responders were younger than non-responders (mean age 55·6 years [SD 9·9] vs 60·4 years [8·6]; p<0·0001). The DPP validation cohort included 683 participants who lost at least 5% of their bodyweight at baseline. Of these, 132 (19%) were responders and 551 (81%) were non-responders. In PLIS, BMI reduction was similar between responders and non-responders (responders mean at baseline 32·4 kg/m 2 [SD 5·6] to mean at 12 months 29·0 kg/m 2 [4·9] vs non-responders 32·1 kg/m 2 [5·9] to 29·2 kg/m 2 [5·4]; p=0·86). However, whole-body insulin sensitivity increased more in responders than in non-responders (mean at baseline 291 mL/[min × m 2 ], SD 60 to mean at 12 months 378 mL/[min × m 2 ], 56 vs 278 mL/[min × m 2 ], 62, to 323 mL/[min × m 2 ], 66; p<0·0001), whereas insulin secretion did not differ within groups over time or between groups (responders mean at baseline 175 pmol/mmol [SD 64] to mean at 12 months 163·7 pmol/mmol [60·6] vs non-responders 158·0 pmol/mmol [55·6] to 154·1 pmol/mmol [56·2]; p=0·46). IHL decreased in both groups, without a difference between groups (responders mean at baseline 10·1% [SD 8·7] to mean at 12 months 3·5% [3·9] vs non-responders 10·3% [8·1] to 4·2% [4·2]; p=0·34); however, VAT decreased more in responders than in non-responders (mean at baseline 6·2 L [SD 2·9] to mean at 12 months 4·1 L [2·3] vs 5·7 L [2·3] to 4·5 L [2·2]; p=0·0003). Responders had a 73% lower risk of developing type 2 diabetes than non-responders in the 2 years after the intervention ended., Interpretation: By contrast to remission of type 2 diabetes, resolution of prediabetes was characterised by an improvement in insulin sensitivity and reduced VAT. Because return to normal glucose regulation (NGR) prevents development of type 2 diabetes, we propose the concept of remission of prediabetes in analogy to type 2 diabetes. We suggest that remission of prediabetes should be the primary therapeutic aim in individuals with prediabetes., Funding: German Federal Ministry for Education and Research via the German Center for Diabetes Research; the Ministry of Science, Research and the Arts Baden-Württemberg; the Helmholtz Association and Helmholtz Munich; the Cluster of Excellence Controlling Microbes to Fight Infections; and the German Research Foundation., Competing Interests: Declaration of interests SKab receives grants from the German Diabetes Association, Wilhelm-Doerenkamp-Foundation, and California Walnut Commission and lecture fees from Sanofi, Boehringer Ingelheim, Berlin-Chemie, Lilly Deutschland, and Juzo Akademie. RW receives lecture fees from Eli Lilly, Novo Nordisk, and Sanofi and participates in a Data Safety Monitoring Board and Advisory Board for Eli Lilly. AL receives grants from Helmholtz Munich and Ludwig Maximilian University Klinikum München. NP receives lecture fees from NovoNordisk and Gesellschaft für Wissenschaftstransfer Dresden; advisory board fees from Bayer Vital; and holds patent WO 2021/092265A1 for diagnosis and treatment of non-alcoholic fatty liver disease and liver fibrosis. VN is the founder and equity owner of sThesis, iThera Medical, Spear UG, and I3. All other authors declare no competing interests., (Copyright © 2023 Elsevier Ltd. All rights reserved.)