Antidiabetic Medication and Asthma Attacks.

Importance: Elevated body mass index (BMI) and type 2 diabetes are prevalent in asthma and are associated with an increase in the risk of asthma attacks. In experimental studies, the diabetes medications metformin and glucagon-like peptide-1 receptor agonists (GLP-1RA) have mitigated airway inflammation, hyperresponsiveness, and remodeling. However, epidemiological evidence is limited. Objective: To estimate the association of metformin and add-on antidiabetic medications (GLP-1RA, dipeptidyl peptidase-4 inhibitors, sulphonylureas, sodium-glucose cotransporter-2 inhibitors, and insulin) with asthma attacks. Design, Setting, and Participants: The study used data from the UK Clinical Practice Research Datalink (CPRD) Aurum linked hospital admissions and mortality data from 2004 to 2020. A triangulation approach was used that applied 2 distinct approaches to enhance robustness: a self-controlled case series (SCCS) and a metformin new user cohort with inverse probability of treatment weighting (IPTW). Eligible participants were new users of metformin with type 2 diabetes. To evaluate the association between metabolic phenotypes (BMI, glycemic control) and asthma phenotypes (type 2 inflammation, asthma severity), interaction analyses were conducted. Negative control analyses were conducted to assess for bias. Exposure: The primary exposure was metformin; secondary exposures included add-on antidiabetic medications. Main Outcomes: The primary outcome was first asthma exacerbation (short course of oral corticosteroids, unscheduled asthma-related hospital attendance, or death) during 12-month follow-up. Incidence rate ratios (IRRs) with 95% CIs were estimated using fixed-effect conditional Poisson models in the SCCS, and hazard ratios (HRs) were estimated using weighted Cox proportional hazards models in the cohort. Results: Of more than 2 million adults with asthma, 4278 patients (2617 women [61.2%]; mean [SD] age, 52.9 [13.6] years) were identified for the SCCS and 8424 patients (4690 women [55.7%]; unexposed: mean [SD] age, 61.6 [13.2] years; exposed: mean [SD] age, 59.7 [13.7] years) for the IPTW cohort. Metformin was found to be associated with fewer asthma attacks of similar magnitude in both approaches (SCCS: IRR, 0.68; 95% CI, 0.62-0.75; IPTW: HR, 0.76; 95% CI, 0.67-0.85). Negative control analyses did not find evidence of significant bias. Hemoglobin A1c levels, BMI, blood eosinophil cell counts, and asthma severity did not modify the association. The only add-on antidiabetic medication to have an additive association was GLP-1RA (SCCS: IRR, 0.60; 95% CI, 0.49-0.73). Conclusions and Relevance: The results of this cohort study suggest that metformin was associated with a lower rate of asthma attacks, with further reductions with the use of GLP-1RA. This appeared to be associated with mechanisms other than through glycemic control or weight loss and occurred across asthma phenotypes.