OBJECTIVE Data regarding the effects of sodium–glucose cotransporter 2 inhibitors in the elderly (age ≥65 years) and very elderly (age ≥75 years) are limited. RESEARCH DESIGN AND METHODS The Dapagliflozin Effect on Cardiovascular Events (DECLARE)–TIMI 58 assessed cardiac and renal outcomes of dapagliflozin versus placebo in patients with type 2 diabetes. Efficacy and safety outcomes were studied within age subgroups for treatment effect and age-based treatment interaction. RESULTS Of the 17,160 patients, 9,253 were <65 years of age, 6,811 ≥65 to <75 years, and 1,096 ≥75 years. Dapagliflozin reduced the composite of cardiovascular death or hospitalization for heart failure consistently, with a hazard ratio (HR) of 0.88 (95% CI 0.72, 1.07), 0.77 (0.63, 0.94), and 0.94 (0.65, 1.36) in age-groups <65, ≥65 to <75, and ≥75 years, respectively (interaction P value 0.5277). Overall, dapagliflozin did not significantly decrease the rates of major adverse cardiovascular events, with HR 0.93 (95% CI 0.81, 1.08), 0.97 (0.83, 1.13), and 0.84 (0.61, 1.15) in age-groups <65, ≥65 to <75, and ≥75 years, respectively (interaction P value 0.7352). The relative risk reduction for the secondary prespecified cardiorenal composite outcome ranged from 18% to 28% in the different age-groups with no heterogeneity. Major hypoglycemia was less frequent with dapagliflozin versus placebo, with HR 0.97 (95% CI 0.58, 1.64), 0.50 (0.29, 0.84), and 0.68 (0.29, 1.57) in age-groups <65, ≥65 to <75, and ≥75 years, respectively (interaction P value 0.2107). Safety outcomes, including fractures, volume depletion, cancer, urinary tract infections, and amputations were balanced with dapagliflozin versus placebo, and acute kidney injury was reduced, all regardless of age. Genital infections that were serious or led to discontinuation of the study drug and diabetic ketoacidosis were uncommon, yet more frequent with dapagliflozin versus placebo, without heterogeneity (interaction P values 0.1058 and 0.8433, respectively). CONCLUSIONS The overall efficacy and safety of dapagliflozin are consistent regardless of age.

OBJECTIVE: Sodium–glucose cotransporter 2 inhibitors (SGLT2i) improve albuminuria in patients with high cardiorenal risk. We report albuminuria change in the Dapagliflozin Effect on Cardiovascular Events (DECLARE-TIMI 58) cardiovascular outcome trial, which included populations with lower cardiorenal risk. RESEARCH DESIGN AND METHODS: DECLARE-TIMI 58 randomized 17,160 patients with type 2 diabetes, creatinine clearance >60 mL/min, and either atherosclerotic cardiovascular disease (CVD; 40.6%) or risk-factors for CVD (59.4%) to dapagliflozin or placebo. Urinary albumin-to-creatinine ratio (UACR) was tested at baseline, 6 months, 12 months, and yearly thereafter. The change in UACR over time was measured as a continuous and categorical variable (≤15, >15 to 300 mg/g) by treatment arm. The composite cardiorenal outcome was a ≥40% sustained decline in the estimated glomerular filtration rate (eGFR) to 15 to 300 mg/g. Measured as a continuous variable, UACR improved from baseline to 4.0 years with dapagliflozin, compared with placebo, across all UACR and eGFR categories (all P < 0.0001). Sustained confirmed ≥1 category improvement in UACR was more common in dapagliflozin versus placebo (hazard ratio 1.45 [95% CI 1.35–1.56], P < 0.0001). Cardiorenal outcome was reduced with dapagliflozin for subgroups of UACR ≥30 mg/g (P < 0.0125, Pinteraction = 0.033), and the renal-specific outcome was reduced for all UACR subgroups (P < 0.05, Pinteraction = 0.480). CONCLUSIONS: In DECLARE-TIMI 58, dapagliflozin demonstrated a favorable effect on UACR and renal-specific outcome across baseline UACR categories, including patients with normal albumin excretion. The results suggest a role for SGLT2i also in the primary prevention of diabetic kidney disease.

Aims: Identification, a priori, of those at high risk of progression from pre‐diabetes to diabetes may enable targeted delivery of interventional programmes while avoiding the burden of prevention and treatment in those at low risk. We studied whether the use of a machine‐learning model can improve the prediction of incident diabetes utilizing patient data from electronic medical records.Methods: A machine‐learning model predicting the progression from pre‐diabetes to diabetes was developed using a gradient boosted trees model. The model was trained on data from The Health Improvement Network (THIN) database cohort, internally validated on THIN data not used for training, and externally validated on the Canadian AppleTree and the Israeli Maccabi Health Services (MHS) data sets. The model's predictive ability was compared with that of a logistic‐regression model within each data set.Results: A cohort of 852 454 individuals with pre‐diabetes (glucose ≥ 100 mg/dL and/or HbA1c ≥ 5.7) was used for model training including 4.9 million time points using 900 features. The full model was eventually implemented using 69 variables, generated from 11 basic signals. The machine‐learning model demonstrated superiority over the logistic‐regression model, which was maintained at all sensitivity levels – comparing AUC [95% CI] between the models; in the THIN data set (0.865 [0.860,0.869] vs 0.778 [0.773,0.784] P < .05), the AppleTree data set (0.907 [0.896, 0.919] vs 0.880 [0.867, 0.894] P < .05) and the MHS data set (0.925 [0.923, 0.927] vs 0.876 [0.872, 0.879] P < .05).Conclusions: Machine‐learning models preserve their performance across populations in diabetes prediction, and can be integrated into large clinical systems, leading to judicious selection of persons for interventional programmes.

Type 2 diabetes is a progressive disorder characterized by increasing hyperglycemia and the need to gradually intensify therapy in order to achieve and maintain glycemic control. Early initiation of combination therapy has been proposed as an approach to achieve glycemic goals earlier and delay the deterioration of glycemic control and with possible better preservation of β-cell function. We discuss in this article the pros and cons of this approach, focusing on individuals with HbA1c at diagnosis of 7.5–9.0%, where difference of opinion still exists on management. Initial combination therapy is proposed to lead to better and faster achievement of glycemic targets versus monotherapy and to impede clinical inertia and may possibly slow the deterioration of β-cell function. However, treating patients with sequential therapy is proposed to allow one to fully assess the efficacy and risk-to-benefit ratio of each drug as it is added. Furthermore, there is no evidence to support that rapid addition and titration of medications according to the glycemic profile achieved are inferior to initial combination therapy if glycemic targets are attained in a timely manner. Initial combination therapy is argued to postpone clinical inertia to the next decision point but does not eliminate it. Additionally, it may have been the agents chosen and not the timing of their initiation that led to improved β-cell function in the studies of initial combination therapy, and there are no data currently comparing use of the same drugs initiated simultaneously or sequentially. Heightened awareness of providers, individualization of therapy and setting, and reaching glycemic targets remain the mainstays of care.