S. E. Holden et al. Diabetes, Obesity and Metabolism. Doi: 10.1111/dom.12412
The authors present another pharmacoepidemiological study in type 2 diabetes. To the uninitiated, this means they accessed routinely collected data and examined for associations between outcomes (case all-cause mortality, cardiovascular events, cancer rates etc.) and prescribed medication. 6,484 subjects from the UK Clinical Practice Research Datalink (CPRD) who progressed to insulin monotherapy from year 2000 were ‘followed’ for 3.3 years. ‘Followed’ is in parentheses since these analyses are retrospective; they also cannot tease out cause-and-effect. The headline was an association between increasing insulin dose and adverse outcomes, implying that insulin is harmful. This comes from a group with a long-standing anti-insulin agenda; indeed, one might question what this adds to their 2013 publication which reached the same conclusion (also from CPRD). Is it relevant to the typical person with type 2 diabetes in the UK? I think not, since it is highly unusual for patients to be treated only with insulin. Will it affect clinical practice? I doubt it, since insulin is currently used as the treatment of last resort when options have been exhausted.