The Risk for Type 2 Diabetes with Inhaled Corticosteroids

To what extent can the increased likelihood of Type 2 diabetes in patients with COPD be attributed to the use of inhaled corticosteroids?  

Prospective, randomized trials have not detected an association between the use of inhaled steroids for COPD and new-onset diabetes mellitus or an increase in its severity, but such trials may have been too small to detect an effect. Suissa and colleagues used a large Canadian pharmacy database, which provided data on all prescriptions and medical service usage of 3 million enrollees. 

The investigators followed 388,584 patients who were being treated with COPD medications (and not on T2DM medications at study entry) for a mean of 5.5 years. During that period, 30,167 of the patients with COPD began treatment with antidiabetic drugs, an incidence rate of 14.2 new diabetes diagnoses per 1,000 patients with COPD per year. For comparison, they also followed 301,096 age-matched patients. They found that the use of inhaled steroids was associated with a 34% mean increase in the incidence of new-onset T2DM, a statistically significant increase. Furthermore, the frequency of developing T2DM was dose related, with higher doses being associated with a greater likelihood of new-onset T2DM. 

Oral corticosteroid use is well known to carry the risk for glucose intolerance and T2DM, a risk that is clearly dose related. Although previously sought, a corresponding risk associated with the use of inhaled steroids has never been established. In light of the fact that more than 70% of patients with COPD currently use an inhaled corticosteroid and recent findings that inhaled corticosteroid use by patients with COPD is associated with a small but significant increase in pneumonia, the current report presents physicians and patients with new concerns. How strong is the evidence for a cause-and-effect relationship with new-onset T2DM, and how great is the risk? 

The gold standard for treatment outcomes is the prospective, randomized controlled trial, but such a study would need to be impossibly large to answer the present question. However, observational studies such as this one present numerous problems, the biggest one being that it is difficult — if not impossible — to match the study cohort with appropriate controls and to adjust for the many confounding factors. For instance, pharmacy databases do not contain relevant information, such as body mass index. The demographic data in the present report also show possibly important differences between cases and controls, with higher frequencies of oral steroid use and comorbidities among the cases. 

However, let us assume that there is a real risk. How large is it, and should it affect our treatment of COPD? The highest dose of inhaled corticosteroid currently approved for COPD in the United States (unlike in Canada) is 500 µg fluticasone per day. This would be at the lower end of the relative risk, namely, between 1.18 and 1.34 — a small but still a statistically and clinically significant risk. However, this risk must be weighed against the benefit of corticosteroid use in COPD. Although the latter are not nearly as effective in COPD as in asthma, they have been shown to reduce the frequency of acute exacerbations by up to 25%.  

Although there may be a real risk for the development of T2DM in patients with COPD whose maintenance treatment includes inhaled corticosteroids, the risk is small with daily dosages of 500 µg fluticasone or an equivalent steroid per day or less, and their regular use should be limited to patients who are experiencing acute exacerbations of COPD, as recommended by current guidelines.