According to an editorial, the findings of patients prescribed more antihyperglycaemic agents experiencing less hyperglycaemia and, when combined with sulfonylureas, more hypoglycaemia, are unsurprising, whilst the interpretation of their results in relation to end organ complications is more difficult. It notes that “drug choice tends to track with disease duration and control—metformin is used early and when diabetes is mild, sulfonylureas, gliptins, and glitazones (and other drug classes) are added on, followed by insulin based regimens. Thus, drugs that are used later in the course of disease may be spuriously associated with complications that also occur years after diagnosis. This may explain why in this study insulin use is associated with complications from diabetes. Similarly, factors related to duration of disease and comorbidities may explain why in this study the incidence of renal failure was greater among people using gliptins or glitazones compared with those using metformin only. These alternative explanations reduce confidence in identified associations to the point that clinicians and patients are no further ahead.” The commentator is not aware of any antihyperglycaemic agent that has been shown to convincingly reduce the risk of any diabetes related complication to any greater extent than other agents, though promising data from recent large clinical trials have suggested mortality reductions with sodium-glucose cotransporter-2 inhibitors and prevention of recurrent strokes with pioglitazone. However he notes that “clinicians and patients must remain sceptical, given how the evidence about drugs accrues: only some questions get answers, and only some answers get published. Sadly, when it comes to antidiabetes drugs, reports of exciting benefits and catastrophic harms are often exaggerations.”