The evidence on the risk of diabetes from inhaled corticosteroids (ICS) has been ambivalent so far. A recent large real-world study provides valuable national registry data from primary care.
Do inhaled corticosteroids (ICS) influence the diabetes risk of COPD patients? Previous study results were mixed. Some previous studies have linked high-dose ICS in particular to an increased risk of new disease and worsening of existing diabetes, while other studies have not reported such an association.
Researchers at the University of Uppsala, Sweden, therefore analysed high-quality data from over seven thousand patients.1 As these were from primary care practices, the observations should be representative of the general population and clinical practice.
At the time of the study, the basic situation regarding ICS therapy in Sweden was comparable to that of other countries: national guidelines recommended the use of ICS only in combination with long-acting β2 adrenergic receptor agonists (LABA) for patients with FEV1 (one-second capacity) < 50% and a history of exacerbations. Nevertheless, real-world surveys have shown that many patients in GOLD stages A and B also receive ICSs/ LABA.2 The authors believe that this could indicate that more patients are exposed to the potential risks of ICS therapy than necessary.
In order to focus on the effects of inhaled steroids, patients who had received oral corticosteroids (OCS) more than once in the study period and in the previous year were excluded. This left 7,078 COPD patients with an average age of 68.6 years (55.7% women). The cumulative 5-year incidence rate per 100,000 person-years was 1,507 and the annual incidence rate per 100,000 person-years ranged from 850 to 1,919.
Compared to the reference cohort (without ICS therapy), the relative risk of developing new type 2 diabetes after exposure to ICS increased significantly: by 32% (p = 0.0201) for patients under stable low ICS dose and by 64% (p = 0.0088) for patients under continuous high dose.
High but unstable doses were even associated with a doubling of the risk: the risk associated with a fluctuating use was 96% (p = 0.0234) and a change from stable low to stable high doses was 100% higher (p = 0.0088). In contrast, the time to develop diabetes did not appear to be dose-dependent, and no significant differences were observed between the ICS dose groups.
One of the limitations of the otherwise good and valuable study is the retrospective design. It is necessary to take into account a multitude of other factors that also influence the risk of disease, for which data were certainly only partially available here. However, a correction for hypertension and heart failure was made, which had no significant effect on the results.
The results are consistent with those of many other studies, including a very large cohort study from 2019, which was based on two British databases.3 Long-term therapy with average ICS doses > 500 µg/day was also associated with a significantly increased risk of diabetes (new onset and progression) and osteoporosis. The topic therefore deserves further evaluation.
References:
1. Ställberg, B. et al. Inhaled corticosteroids and the risk of type 2 diabetes among Swedish COPD patients. npj Primary Care Respiratory Medicine 30, 1-3 (2020).
2. Sundh, J. et al. Factors influencing pharmacological treatment in COPD: a comparison of 2005 and 2014 European Clinical Respiratory Journal 4, 1409060 (2017).
3. Price, D. B. et al. Inhaled corticosteroids in COPD and onset of type 2 diabetes and osteoporosis: matched cohort study. npj Primary Care Respiratory Medicine 29, 1-13 (2019).