During the three years of the DPP, ILS reduced incidence of diabetes compared with placebo in high-risk adults. Diabetes was defined using ADA criteria based on 1) annual OGTT or semi-annual fasting glucose (the primary DPP outcome, diab-G, hazard ratio (HR) =0.42 and rate difference (RD) = -6.2 cases/100 person-years) or 2) annual HbA1c ≥ 6.5% (diab‑A1C, HR=0.51 and RD = -4.2) . After DPP ended, participants were unmasked, placebo was discontinued, and all participants were offered group-based ILS and followed in the DPP Outcomes Study. We now compare the original ILS and placebo groups over a mean of 21 years since randomization. We test heterogeneity across subgroups defined by demographics and baseline glucose, HbA1c, BMI, and history of gestational diabetes and express it by interaction p-values (P-het) . During the total follow-up, ILS, compared with placebo, significantly reduced incidence of diab‑G: HR=0.76 (95% CI=0.68, 0.85) , RD= -1.62 (-2.28, -0.97) and of diab-A1C: HR=0.80 (0.70, 0.92) , RD= -1.67 (‑2.24, -1.10) . Effects on diab-G were homogenous across subgroups except for fasting glucose (P-het <0.01) and HbA1c (P-het <0.02) with greater effects at higher baseline values. By contrast, ILS reduced incidence of diab-A1C more in older than younger persons (P-het <0.02) , and more in men than women (P-het <0.03) . Notably, ILS was highly effective in reducing incidence of diab‑A1C if baseline HbA1c was already elevated (HbA1c=6.0-6.4%: HR=0.62, RD= -3.27) but ineffective if baseline HbA1c was normal by ADA criteria (HbA1c <5.7%: HR=1.09, RD= +0.19; P-het <0.001) . In sum, ILS effects on reducing incidence of diab-G and diab-A1C persisted but decreased during total follow-up from that observed during DPP (i.e., HRs and RDs closer to 1 and 0) . ILS effects on diab-G were greater in those with higher baseline fasting glucose or HbA1c. Effects on diab-A1C were greater in men, older participants, and those with higher baseline HbA1c.


W.C. Knowler: None. S. Edelstein: None. P.H. Bennett: Stock/Shareholder; Baxter INTERNTN Health Care, Merck & Co., Inc., UnitedHealth Group. D. Dabelea: None. M.A. Hoskin: None. S.E. Kahn: Advisory Panel; Bayer AG, Boehringer Ingelheim International GmbH, Eli Lilly and Company, Intarcia Therapeutics, Inc., Merck & Co., Inc., Novo Nordisk, Pfizer Inc. R. Kalyani: None. C. Kim: None. S. Raghavan: None. M. Temprosa: None. E.M. Venditti: Advisory Panel; PeopleOne Health. D.M. Nathan: None. D. Research Group: None.


National Institute of Diabetes and Digestive and Kidney Diseases (UDK048489, UDK048339, UDK048377, UDK048349, UDK048381, UDK048468, UDK048434, UDK048485, UDK048375, UDK048514, UDK048437, UDK048413, UDK048411, UDK048406, UDK048380, UDK048397, UDK048412, UDK048404, UDK048387, UDK048407, UDK048443, and UDK048400)

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at http://www.diabetesjournals.org/content/license.