A335and a Bayesian network meta-analysis of log-hazard ratios was performed. The primary outcome of the network meta-analysis was the time to progression to Type 2 diabetes mellitus. Results are presented as hazard ratios and the probabilities of treatment rankings. Results: 30 studies were included in the network meta-analysis. There was a reduced hazard of progression to Type 2 diabetes mellitus associated with all interventions versus standard care. The most effective interventions compared to standard care were diet plus pioglitazone (HR 0.17, 95% CrI [0.09, 0. 33] ), glipizide (HR 0.16, 95% CrI [0.02, 1.62] ), diet plus exercise plus metformin plus rosiglitazone (HR 0.20, 95% CrI [0.11, 0.39] ), diet plus exercise plus orlistat (HR 0.31, 95% CrI [0.16, 0.61] ) and diet plus exercise plus pedometer (HR 0.35 95% CrI [0.11, 1.14] ). The least effective intervention was ramipril (HR 0.91, 95% CrI [0.72, 1.14] ). ConClusions: Pharmacological and lifestyle interventions are beneficial in reducing the risk of progression to Type 2 diabetes mellitus. Lifestyle interventions require significant behaviour changes and this may be achieved through incentives such as the use of pedometers. Lifestyle interventions alone, whilst beneficial, are unlikely to be as effective as pharmacological interventions alone or in combination with lifestyle interventions. Adverse events and costs of pharmacological interventions should be taken into account when considering potential risks and benefits, and their cost-effectiveness relative to lifestyle interventions.objeCtives: Previous meta-analyses show inconsistent results regarding benefits of insulin glargine (GLA) over NPH insulin (NPH) in terms of nocturnal hypoglycaemia. We analysed standardized efficacy and safety outcomes in uncontrolled insulin-naïve subjects with T2DM treated with GLA or NPH according to the oral antidiabetic drug (OAD) to which insulin was added (sulfonylurea [SU] ± metformin [MET] ). Methods: Patient-level data from 4 Treat-To-Target (TTT) RCTs (FPG < 100 mg/dL) of ≥ 24 weeks duration were pooled. HbA 1c , weight, dose, and hypoglycaemia (overall and nocturnal, plasma glucose < 56 mg/dL) were assessed. Results: 2,091 subjects were analysed; 49% male, mean age 57.8 years. SU-treated subjects had longer T2DM duration, higher baseline HbA 1c , and were less obese than MET+SUtreated subjects. Endpoint HbA 1c values were similar for GLA and NPH overall (7.4 vs 7.5%). Subjects adding GLA or NPH to MET+SU had numerically lower weightadjusted endpoint insulin doses (GLA 0.40 vs 0.44 U/kg; NPH 0.36 vs 0.42 U/kg), with less weight gain (GLA: +1.9 vs +3.7 kg; NPH: +1.8 vs +3.0 kg) versus subjects adding insulin to SU. Hypoglycaemia incidence (overall: odds ratio [OR] 0.79 [95% CI 0.66−0.95]; P= 0.011, nocturnal: OR 0.64 [0.51−0.81]; P< 0.001) and event rates (overall: rate ratio [RR] 0.78 [0.65−0.93]; P= 0.006, nocturnal: RR 0.54 [0.42−0.69]; P< 0.001)were significantly lower in GLA-treated subjects versus NPH, irrespective of concomitant OAD. In general, h...