Association of Metformin Use with Asthma Exacerbation in Patients with Concurrent Asthma and Diabetes: A Systematic Review and Meta-Analysis of Observational Studies

Background Asthma and diabetes are both diseases that affect a wide range of people worldwide. As a common treatment for diabetes, metformin has also been reported to be effective in improving asthma outcomes. We conducted a combined analysis to examine the efficacy of metformin in reducing asthma exacerbation in patients with concurrent asthma and diabetes. Methods We searched the PubMed, Embase, and CENTRAL databases for articles published prior to April 2020 to find observational studies of individuals with concurrent asthma and diabetes that compared the risk of asthma exacerbation between metformin users and nonusers. Two researchers separately screened the studies, extracted data, and evaluated the risk of bias. The primary outcome was the adjusted risk of asthma exacerbation. The secondary outcomes were the adjusted risk of asthma-related hospitalization and emergency room visits. Review Manager was used for data analysis and plotting. I2 and χ2 tests were used to estimate heterogeneity. A random effects or fixed effects model was used depending on the heterogeneity. Odds ratios were calculated for dichotomous variables. Results We included two studies with a total of 25252 patients. The pooled effect size showed that metformin was inversely associated with a risk of asthma exacerbation (OR = 0.65, 95% CI 0.28–1.48; χ2 = 5.42, P=0.02; I2 = 82%), asthma-related emergency department visits (OR = 0.81, 95% CI 0.74–0.89; χ2 = 0.36, P=0.55; I2 = 0%), and hospitalizations (OR = 0.43, 95% CI 0.14–1.29; χ2 = 4.01, P=0.05; I2 = 75%). Conclusion This meta-analysis suggested that metformin decreased the risk of asthma-related emergency room visits for patients with concurrent asthma and diabetes. Metformin reduced the risk of asthma-related hospitalization and exacerbation but was not statistically significant. More randomized trials involving larger samples should be considered, and the mechanisms of these effects need to be fully elucidated.


Introduction
Asthma, a chronic inflammatory disease of the small airway characterized by recurring wheezing, reversible airflow obstruction, cough, and other symptoms [1], affects more than 300 million people worldwide [2]. e WHO estimates that 235 million people currently suffer from asthma [3]. In China, more than 40 million people are affected by asthma, and the number has been increasing rapidly [4]. Asthma is treatable, but there is no cure [3]. It is estimated that more than 50% of asthma patients do not have control of the disease. Life-threatening exacerbations affect morbidity and mortality [5]. Controller and reliever medications are the main treatments for asthma [6]. e use of single maintenance and reliever therapy (SMART) was reported to be an excellent strategy for decreasing the risk of asthma exacerbations [6]. In addition, a series of monoclonal antibodies emerged against conventional drugs [7,8]. However, the efficacy and safety of these new agents are not conclusive, and their price is not affordable for everyone. erefore, cheap and effective adjuvant drugs are desirable. Diabetes mellitus (DM) is a group of metabolic disorders characterized by a high blood sugar level, and DM is a common comorbid condition among adult patients with asthma [9]. Recent epidemiological studies have shown that the incidence rate of asthma in patients with type 2 diabetes is low, as metformin reduces the risk of asthma in patients with type 2 DM [9,10]. In patients with concurrent asthma and diabetes population, the risk of asthma exacerbation is lower for metformin users [11,12]. However, the efficacy of metformin for treating asthma has not been conclusive due to a lack of studies. We performed a systematic review and meta-analysis on this subject. Conclusions from such an analysis may help determine whether to use metformin as an adjuvant therapy for asthma.

Protocol and Registration.
e protocol of the present systematic review was registered in the international platform of registered systematic review and meta-analysis protocols (INPLASY, https://inplasy.com/) and was reported following PRISMA [13]

Study Selection.
e inclusion criteria were as follows: (1) observational studies; (2) patients age>18 years; (3) examined the use of metformin in patients with concurrent asthma and diabetes; (4) the odds ratios (ORs) or hazard ratios (HRs) of asthma exacerbation were compared between metformin users and nonusers. e exclusion criteria were as follows: (1) animal or cell research; (2) reviews, letters to the editor, or case reports; (3) duplicate articles.
Two authors independently checked the titles and abstracts (GJ and GL). If there were different opinions between the reviewers, another author (ZM) was consulted to reach a consensus as the third investigator.

Data Collection
Process. Data were extracted from each selected study (LH and YH), including the name of the author, year of publication, geographical location, study design, case number of metformin user/nonuser, age of metformin user/nonuser, gender of metformin user/nonuser, inclusion and exclusion criteria, follow-up, outcomes, and statistical approaches. We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [13]. e quality of individual records was assessed according to the Newcastle-Ottawa Scale [14].

Outcomes.
e primary outcome was the adjusted risk of asthma exacerbation. e secondary outcomes were the adjusted risk of asthma-related hospitalization and emergency room visits.

Statistical Analysis.
Review Manager software (version 5.3; Cochrane Collaboration, Oxford, UK) was used to estimate the risks of bias of the included studies, analyze data, and create plots (WQ, YC, and CY). e sensitivity analysis and publication bias test were performed using R language if enough original studies were included. I 2 and χ 2 tests were used to estimate heterogeneity. If P > 0.1 or I 2 <40%, a fixed effects model was used for the analysis. If there was a high degree of heterogeneity, a random effects analysis was used. Odds ratios (ORs) were calculated for dichotomous variables.

Study Description and Risk of Bias.
By carrying out the search strategy mentioned above, a total of 106 articles were found after duplicated records were removed. After the title and the abstracts were screened, we downloaded the full texts of nine records, of which two were ultimately included in our analysis, including a total of 25252 participants. e details of the study selection process are shown in Figure 1. In total, two observational studies were included in the present meta-analysis, and the characteristics of the studies are shown in Table 1.
e two trials were high-quality studies. e Newcastle-Ottawa Scale was used to evaluate the risk of bias for the observational studies (Table 2).
Two observational studies reported the risk of exacerbation as the main outcome. e pooled effect showed that metformin decreased the risk of asthma exacerbation in patients with concurrent asthma and diabetes. ere was a high level of heterogeneity, and the effect of metformin was not significant (OR � 0.65, 95% CI 0.28-1.48; χ 2 � 5.42, P � 0.02; I 2 � 82%; Figure 2).

Discussion
In this meta-analysis of adults with asthma and type 2 DM, metformin use was inversely associated with a risk of asthma exacerbation, asthma-related emergency department visits, and hospitalizations, although the findings for the exacerbation and hospitalization were not statistically significant.
If we focus on the studies individually, metformin was shown to be associated with a decreased risk of the three outcomes in both studies. However, the nonsignificant results might be due to the heterogeneity. Both of the included studies analyzed overly heterogeneous population, with different average age points, inclusion and exclusion criteria, follow-up periods, and statistical methods. On the one hand, this made it impossible to draw definitive conclusions regarding the subject matter. As shown above, there was a high level of heterogeneity and the effect of metformin was not   significant. On the other hand, the two research studies validated each other with different populations. To some extent, they are likely to have promoted the generalizability of the effect of metformin on asthma with type 2 DM. We expected that more homogeneous studies will soon emerge in the future, so that researchers could draw more definitive conclusions. Type 2 DM and asthma are both extremely heterogeneous diseases in terms of pathogenesis and clinical characteristics. Metabolismrelated asthma is not the same as allergic asthma, and eosinophilic and neutrophilic asthma may differ in the response of metformin. It is regrettable that the included two studies did not report the subgroup data of the phenotypes regarding asthma or type 2 DM.

Canadian Respiratory Journal
Drug therapies used in the group that did not take metformin were mentioned in the baseline characteristics of the two cohorts. In one article [12], there were no difference of the percentage of drugs use between two groups for asthma and type 2 DM. However, in another publication [11], there were different proportions of SABA, corticosteroid, methylxanthine, meglitinide, and insulin users between observational and control groups. Insulin has been reported to be associated with increased risk of asthma [10]. Different corticosteroid user percentages indicated that severity between two groups may not be the same. is could represent one of the methodological biases that account for the nonsignificance of the results and the heterogeneity. However, these findings suggest a potential application for metformin in decreasing the risk of severe asthma exacerbations.
Metformin, a widely used medication for type 2 DM, has been shown to be associated with a decreased risk of chronic lower respiratory disease mortality, and no similar effect has been found for other antidiabetic medications [15]. Several studies have reported that metformin use reduces the risk of asthma in patients with diabetes [9,10], but the mechanism of this effect is not fully understood. Several possible mechanisms are discussed as follows.
Obesity, metabolic abnormalities, type 2 DM, and asthma are linked to each other. Obesity is a major risk factor for both type 2 DM and asthma [16]. Obese patients with asthma tend to have more symptoms and exacerbation frequency, as well as decreased response to asthma treatments [17]. Metabolic abnormalities, especially insulin resistance, frequently coexist with obesity, which have been proved to be associated with asthma and pulmonary function decrease [18,19]. Insulin resistance is likely to play an important role in the pathophysiology of asthma [20].
ere is more significant association between obesity and asthma if insulin resistance exists [21]. Fortunately, metformin is an effective agent in improving metabolism, weight loss, and glucose reduction by alleviating insulin resistance. Insulin resistance is the common point in the problems listed above, which would be alleviated effectively by metformin.
Metformin is also associated with changes in many asthma-related cells and cytokines. Ig-E and aryl hydrocarbon receptor-mediated mast cell activation could be inhibited by metformin in vitro and in vivo [22]. Metformin downregulated the expression levels of inflammatory cytokines (IL-1β, IL-4, and IL-6 and TNF-α) in intracerebral hemorrhage model rats [23]. Treatment with metformin Odds ratio IV, random, 95% CI  Odds ratio IV, random, 95% CI Figure 4: Metformin decreased the risk of asthma-related hospitalization in patients with concurrent asthma and diabetes, but this decrease was not statistically significant. 6 Canadian Respiratory Journal showed ability to attenuate upregulation of IL-4-DUOX2 pathway and other pathological damages to the lung after exposure to a high dose of ionizing radiation [24]. In an animal experiment, metformin increased the ratio of Treg/ 17 cells in mice [25]. As we all know, molecules such as IL-4, IL-6, TNF-α, Ig-E, and 17cell played key roles in the pathological process of asthma. Effects of these molecules and cells may be the vital mechanisms of metformin on asthma.
AMP-activated protein kinase (AMPK) is an enzyme that plays an important role in the regulation of insulin signaling and glucose metabolism [26], which could reduce inflammatory responses and airway remodeling in the respiratory system [27,28]. Metformin activates AMPK, which suppresses glycolysis in immune cells and suppresses cytokine production in vitro and in vivo [29]. Metformin has been shown to attenuate allergic eosinophilic airway inflammation in obese mice and restore levels of AMPK in lung tissue [30]. It may also inhibit airway smooth muscle cell proliferation through AMPK-dependent pathways [31]. More than one article reported that metformin reduces airway inflammation, fibrosis, and remodeling by activation of AMP-activated protein kinases and inhibition of mTOR signaling [28,32]. It seems that AMPK is the common ground of asthma and diabetes, which could be activated by metformin.
Other possible mechanisms were also mentioned in the literature. It was found that the p38MAPK signal transduction pathway increased in both type 1 and type 2 diabetes [33], which has been shown to induce an anti-inflammatory response in vascular smooth muscle cells in diabetic mice [34]. Metformin, as an inhibitor of p38MAPK, may act as a potential treatment for asthma. LPS-induced bronchial epithelial cell injury could be ameliorated by metformin via NF-κB signaling suppression [35]. Metformin induced decreases in the VEGFa level in the airway in obese mice to alleviate airway remodeling [36]. ese examples may be mechanisms by which metformin interferes with asthma. In summary, the discovery of the mechanisms of metformin on respiratory disease is limited, and more basic research should be carried out.
Several limitations should be mentioned. Only two studies were ultimately included in this meta-analysis, which is a low number. erefore, subgroup analysis, sensitivity analysis, and publication bias evaluation could not be conducted. In the absence of high-quality randomized controlled trials, the reliability of the conclusions of this review should be verified in the future.
Finally, we expect that clinical trials assessing the use of metformin in patients with asthma with or without DM should be performed.

Conclusion
In summary, this meta-analysis suggested that metformin decreased the risk of asthma-related emergency room visits for patients with concurrent asthma and diabetes. Metformin reduced the risk of asthma-related hospitalization and exacerbation, but these findings were not statistically significant.
ese findings suggest a potential role for metformin in the respiratory system. Prospective cohort or randomized trials should be considered, and the underlying mechanisms of these effects need to be fully elucidated.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this paper.