Long-term weight loss with metformin or lifestyle intervention in the Diabetes prevention Program outcomes study


John W. Apolzan, Pennington Biomedical Research Center
Elizabeth M. Venditti, UPMC Western Psychiatric Hospital
Sharon L. Edelstein, The George Washington University
William C. Knowler, National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK)
Dana Dabelea, Colorado School of Public Health
Edward J. Boyko, VA Puget Sound Health Care System
Xavier Pi-Sunyer, Vagelos College of Physicians and Surgeons
Rita R. Kalyani, Johns Hopkins University
Paul W. Franks, Skånes universitetssjukhus
Preethi Srikanthan, University of California, Los Angeles
Kishore M. Gadde, Pennington Biomedical Research Center
George A. Bray, Pennington Biomedical Research Center
Annie Chatellier, Pennington Biomedical Research Center
Jennifer Arceneaux, Pennington Biomedical Research Center
Amber Dragg, Pennington Biomedical Research Center
Crystal Duncan, Pennington Biomedical Research Center
Frank L. Greenway, Pennington Biomedical Research Center
Erma Levy, Pennington Biomedical Research Center
Monica Lockett, Pennington Biomedical Research Center
Donna H. Ryan, Pennington Biomedical Research Center
David Ehrmann, The University of Chicago
Margaret J. Matulik, The University of Chicago
Kirsten Czech, The University of Chicago
Catherine DeSandre, The University of Chicago
Barry J. Goldstein, Jefferson Medical College
Kevin Furlong, Jefferson Medical College
Kellie A. Smith, Jefferson Medical College
Wendi Wildman, Jefferson Medical College
Constance Pepe, Jefferson Medical College
Ronald B. Goldberg, University of Miami
Jeanette Calles, University of Miami
Juliet Ojito, University of Miami

Document Type

Journal Article

Publication Date



Annals of Internal Medicine








© 2019 American College of Physicians Background: Identifying reliable predictors of long-term weight loss (LTWL) could lead to improved weight management. Objective: To identify some predictors of LTWL. Design: The DPP (Diabetes Prevention Program) was a randomized controlled trial that compared weight loss with metformin, intensive lifestyle intervention (ILS), or placebo. Its Outcomes Study (DPPOS) observed patients after the masked treatment phase ended. (ClinicalTrials.gov: NCT00004992 and NCT00038727) Setting: 27 DPP and DPPOS clinics. Participants: Of the 3234 randomly assigned participants, 1066 lost at least 5% of baseline weight in the first year and were followed for 15 years. Measurements: Treatment assignment, personal characteristics, and weight. Results: After 1 year, 289 (28.5%) participants in the metformin group, 640 (62.6%) in the ILS group, and 137 (13.4%) in the placebo group had lost at least 5% of their weight. After the masked treatment phase ended, the mean weight loss relative to baseline that was maintained between years 6 and 15 was 6.2% (95% CI, 5.2% to 7.2%) in the metformin group, 3.7% (CI, 3.1% to 4.4%) in the ILS group, and 2.8% (CI, 1.3% to 4.4%) in the placebo group. Independent predictors of LTWL included greater weight loss in the first year in all groups, older age and continued metformin use in the metformin group, older age and absence of either diabetes or a family history of diabetes in the ILS group, and higher fasting plasma glucose levels at baseline in the placebo group. Limitation: Post hoc analysis; examination of nonrandomized subsets of randomized groups after year 1. Conclusion: Among persons with weight loss of at least 5% after 1 year, those originally randomly assigned to metformin had the greatest loss during years 6 to 15. Older age and the amount of weight initially lost were the most consistent predictors of LTWL maintenance.