Introduction

Respiratory diseases are common disorders in life. Respiratory diseases are increasing in mortality and morbidity globally each year. It is a major contributor to the global health burden [1]. In addition, chronic respiratory disease is one of the top 10 causes of death worldwide [2]. Diabetes is a common comorbidity among patients with respiratory diseases. People with diabetes are at increased risk of developing sleep apnoea syndrome, chronic obstructive pulmonary disease (COPD), asthma, and lung injury. Hyperinsulinemia, hyperglycemia, and chronic pro-inflammatory states can lead to impaired lung function. There is evidence that GLP-1RAs may improve lung function through mechanisms such as reducing inflammation and tracheal hyperresponsiveness [3, 4].

GLP-1RAs are novel anti-hyperglycemic agents commonly used. In addition to cardiovascular benefits, emerging evidence suggests that such drugs may also be associated with respiratory benefits, for example in patients with COPD [5, 6].

Recently, a population-based cohort study (involving 56,243 participants) discovered that GLP-1 receptor agonists reduce the incidence of severe exacerbation of COPD [6]. Compared with sulfonylureas, GLP-1 receptor agonists were associated with a 30% decreased risk of severe exacerbation (3.5 v 5.0 events per 100 person years; HR 0.70, 95% CI 0.49–0.99) and moderate exacerbation (0.63, 0.43–0.94). Besides, in a previous retrospective cohort study, the use of GLP-1RA was found to result in fewer asthma attacks in adult asthmatics with type 2 diabetes [7]. GLP-1R agonists (reference) compared with SGLT-2 inhibitors (incidence rate ratio [IRR], 2.98; 95% confidence interval [CI], 1.30–6.80), DPP-4 inhibitors (IRR, 2.45; 95% CI 1.54–3.89), sulfonylureas (IRR, 1.83; 95% CI 1.20–2.77), and basal insulin (IRR, 2.58; 95% CI 1.72–3.88). Also, it allowed patients to potentially improve from obstructive sleep apnea syndrome and have less aggravation of the chronic lower respiratory disease [8, 9]. The adjusted incidence rate of first chronic lower respiratory disease admission during follow-up was 10.7 and 20.3 per 1000 person-years for GLP-1RA and DPP-4I users, respectively, resulting in an adjusted hazard ratio of 0.52 (95% CI 0.32–0.85). From the perspective of its possible underlying mechanism. The use of GLP-1RA was found to reduce inflammation and promote lung function in a mouse model of obstructive lung disease [10]. In another previous animal model, liraglutide pretreatment reduced acute pulmonary damage brought on by LPS in mice [11]. Furthermore, GLP-1RA has been shown to significantly attenuate LPS-stimulated eosinophil activation [12]. It may have immunomodulatory effects. Currently, the relationship between GLP-1RAs and respiratory diseases is unclear. To determine the link between GLP-1 agonists and the occurrence of respiratory diseases, we conducted a meta-analysis of all RCTs that satisfied the inclusion criteria. Meanwhile, since GLP-1RAs are clinically used for type 2 diabetes and weight loss, and diabetes and obesity are risk factors for the development of respiratory diseases, we included people with type 2 diabetes, obesity, or overweight for the study [5, 13].

Methods

Data sources and searches

We searched PubMed (MEDLINE), EMBASE, and Cochrane Library databases to find literature on GLP-1RA published from database creation to December 24, 2022, and the literature was not restricted by language. The following were the search keywords used: “Glucagon-like peptide 1 receptor agonist”, “liraglutide”, “albiglutide”, “lixisenatide”, “exenatide”, “dulaglutide”, “semaglutide”, “Diabetes Mellitus, Type 2”, “obesity”, “Overweight” and “randomized controlled trial”. Detailed search strategies can be found in the Additional file.

This meta-analysis follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) reporting guidelines statement. The International Registry of Prospective Systematic Reviews has prospectively recorded the meta-procedure. (CRD42023396138).

Study selection

Inclusion criteria: (1) Adult people with type 2 diabetes mellitus, overweight or obese. (2) RCT comparing GLP-1 receptor agonist with placebo or positive control (Insulin or other oral hypoglycemic drugs). (3) The results reported at least one respiratory disease event of interest (all data from ClinicalTrials.gov on respiratory adverse events).

Exclusion criteria: (1) Pregnant women. (2) GLP-1RA in combination with other hypoglycemic agents (e.g. IDegLira). (3) Comparison between different GLP-1RAs (4) Multiple publications from the same RCT (using the most complete or most recently reported data), reviews, case reports, letters, and conference proceedings on GLP-1RA. (4) Studies with incomplete data or full text not available (Table 1).

Table 1 Inclusion and exclusion criteria
Full size table

Outcome measures

Primary outcome: overall respiratory disease prevalence.

Secondary outcomes: 12 subcategories of respiratory diseases: Pulmonary edema , pneumonia, Bronchitis, Pulmonary fibrosis, Dyspnoea, Acute respiratory failure, Pleural effusion, Asthma, COPD, Sleep apnoea syndrome, Pulmonary embolism, Pulmonary hypertension.

Data extraction and quality assessment

By reading the full text of the included literature, two investigators (MX and RX) independently extracted the following data: First author, year of study publication, clinical trial registration number, duration of follow-up, type of GLP-1 RA, control drug, and trial subject characteristics (age, sex, BMI, and glycosylated hemoglobin [HbA1c] levels). Any discrepancies arising from this process were resolved by consensus or by a third reviewer (PL). The Cochrane Risk of Bias tool was used to assess the quality of each RCT included [14].

Statistical analysis

We analyzed dichotomous outcomes using RR and 95% CI calculated from random or fixed effects models. Heterogeneity between studies was assessed using the χ2 test and the I2 statistic. If heterogeneity was low (P > 0 0.05; I2 ≤ 50%), analyses were performed using the Mantel–Haenszel method and fixed-effects models. Otherwise, a random-effects model using the DerSimonian-Laird method was used. We also assessed the association of specific GLP-1 RA drugs with respiratory disease. Also, we performed subgroup analyses according to control type, trial duration, and indication. Sensitivity analyses were performed by omitting eligible trials one by one. We used a funnel plot and the Egger test to check for publication bias. All of the analyses were conducted using RevMan 5.4 and Stata 14. We think it is statistically significant if the P-value is < 0.05.

Result

Study search and study characteristics

A search of the above databases based on keywords returned 9570 records in total. After excluding duplicates, and reading the titles and abstracts, a total of 633 documents were read in full. A total of 28 RCTs involving a total of 77485 participants were finally included for analysis [15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42]. Figure 1 shows the study screening flow chart. In the 28 included trials, the majority included subjects with type 2 diabetes (96.3%) with follow-up periods ranging from 24 weeks to 5.4 years. Among all participants, the average age of the study participants was between 46.0 and 74.2 years. The proportion of males ranged from 19.0 to 69.4%. The average BMI varied from 26.8 to 38.5 kg/m2. Mean glycosylated hemoglobin (HbA1c) ranged from 5.7 to 8.9%. (Table 2).

Fig. 1
figure 1

Process of trial selection

Full size image
Table 2 Baseline characteristics of included studies
Full size table

Risk of bias evaluation

The Cochrane Risk of Bias tool, which provides information on the risk of bias evaluation, was used to evaluate the quality of the included studies (Fig. 2).

Fig. 2
figure 2

Quality assessment (Cochrane risk of bias tool) for included RCTs

Full size image

Main outcome

Incidence of respiratory diseases with all GLP-1 receptor agonists

As shown in Fig. 3, this study included 41,663 participants using GLP-1RAs and 35,822 patients using placebo, oral hypoglycemic agents, or insulin. The GLP-1 RAs group had a lower event rate (3.43%) than the control group (4.22%). Patients with GLP-1RAs had a 14% lower risk of respiratory disease compared to controls (RR 0.86, 95% CI 0.81–0.93; p < 0.0001), with no significant heterogeneity between studies (I2 = 0). In the sensitivity analysis, the pooled results were not significantly altered after repeated omission of each trial. There was no obvious publication bias, according to the funnel plots of this analysis. (Additional file 1: Figure S1, Egger’s test P = 0.859 > 0.05).

Fig. 3
figure 3

Forest plot of GLP-1 receptor agonists versus comparators on the risk of overall Respiratory diseases

Full size image

Incidence of respiratory diseases with different GLP-1 receptor agonists

As shown in Fig. 4, eight trials totaling 13655 patients were included with Semaglutide as the experimental group. Semaglutide reduced the risk of developing respiratory diseases by 18% compared to placebo or other interventions (RR 0.82, 95% CI 0.68–0.97, p = 0.02). In addition, six trials totaling 12,955 patients were included with Liraglutide as the experimental group. Compared to the control group, Liraglutide reduced the risk of respiratory disease by 14% (RR 0.86, 95% CI 0.75–0.98, p = 0.03). Four RCT studies including 12,480 patients provided information on the use of Dulaglutide and the risk of respiratory disease. The results showed that Dulaglutide reduced the risk of respiratory disease by 18% compared to controls (RR 0.82, 95% CI 0.70–0.96, p = 0.02).

Fig. 4
figure 4

Forest plot of specific GLP-1 receptor agonists versus comparators on the risk of overall Respiratory diseases

Full size image

However, other GLP-1 receptor agonists did not affect the overall respiratory disease. The RR for the occurrence of respiratory disease in patients receiving Albiglutide was 0.93 (95% CI 0.79–1.10, p = 0.40) compared to other interventions. The RR for the occurrence of respiratory disease in patients receiving Exenatide was 0.93 (95% CI 0.74–1.18, p = 0.55). The RR for the occurrence of respiratory disease in patients receiving Lixisenatide was 0.83 (95% CI 0.62–1.12, p = 0.22). The RR for the occurrence of respiratory disease in patients receiving Efpeglenatide was 0.76 (95% CI 0.46–1.24, p = 0.27). No inter-study heterogeneity was observed in all of the above analyses.

Subgroup analyses

Based on subgroup analysis of trial duration, control type, and indication. The results showed that trial duration, control type, and indication had no significant effect on the effect of GLP-1 receptor agonists on overall respiratory disease (all P subgroups > 0.05; Fig. 5).

Fig. 5
figure 5

Subgroup analysis: GLP-1 receptor agonists and Respiratory disease. CVOT: Cardiovascular Outcomes Trial

Full size image

Secondary outcome

Random assignment to GLP-1 RA treatment resulted in a reduced risk of Pulmonary edema (RR 0.66, 95% CI 0.44–0.98, p = 0.04; I2 = 11; Additional file 1: Fig S3), Bronchitis (RR 0.86, 95% CI 0.74–1.00, p = 0.04; I2 = 0; Additional file 1: Fig S4). Meanwhile, GLP-1RAs showed a trend towards a reduced risk for nine respiratory diseases, although not reaching statistical significance, including pneumonia (RR 0.89, 95% CI 0.78–1.00, p = 0.05; I2 = 0; Additional file 1: Fig S5), Pulmonary fibrosis (RR 0.39, 95% CI 0.13–1.16, p = 0.09; I2 = 0; Additional file 1: Fig. S6), Dyspnoea (RR 0.78, 95% CI 0.58–1.04, p = 0.09; I2 = 0; Additional file 1: Fig. S7), Acute respiratory failure (RR 0.81, 95% CI 0.57–1.13, p = 0.21; I2 = 0; Additional file 1: Fig. S8), Pleural effusion (RR 0.73, 95% CI 0.48–1.11, p = 0.15; I2 = 6; Additional file 1: Figure S9), Asthma (RR 0.85, 95% CI 0.59–1.24, p = 0.40; I2 = 0; Fig S10), and COPD (RR 0.93. 95% CI 0.76–1.13, p = 0.47; I2 = 0; Additional file 1: Fig S11), Sleep apnoea syndrome (RR 0.81, 95% CI 0.48–1.36, p = 0.43; I2 = 0; Additional file 1: Fig S12), Pulmonary embolism (RR 0.95, 95% CI 0.69–1.31, p = 0.76; I2 = 0; Additional file 1: Fig S13). GLP-1RAs did not affect Pulmonary hypertension (RR 1.02, 95% CI 0.51–2.05, p = 0.95; I2 = 5; Additional file 1: Fig S14). Heterogeneity was absent or low for all meta-analyses analyzed above.

Discussion

The relationship between GLP-1RA and respiratory disease has rarely been studied in clinical studies. In 2021, the results of a meta-analysis that included 7 RCT studies showed that GLP-RA uses resulted in a decreased risk of respiratory disease [43]. This meta-analysis included large RCTs with cardiovascular or renal outcomes as experimental endpoints in patients with type 2 diabetes. Our meta-analysis mainly focuses on people with type 2 diabetes, obesity, or overweight. We expand the crowd even further. Furthermore, it is to determine whether the risk may differ by GLP-1RA type, trial duration, control type, and indication. In this regard, we have further discussed and systematically elaborated.

This meta-analysis examined the connection between the injection of GLP-1 receptor agonists and the incidence of diverse respiratory diseases. Two main discoveries were discovered. Firstly, compared to placebo or other therapies, GLP-1 RAs greatly decreased the incidence of total respiratory illness by 14%, particularly with Semaglutid, Liraglutide, and Dulaglutide. In the forest plot, several studies including Harmony Outlets (2018), LEADER (2016), REWIND (2019), and SUSTAIN-6 (2016) appear to show a low risk of respiratory diseases. These studies are all Cardiovascular Outcomes Trials, which are large multi-country and multi-center RCTs. The weight of these studies is between 10.1 and 23.4%, and the total number of participants is between 3297 and 9901. Median follow-up time of these studies ranged from 1.6 to 5.4 years. Therefore, more participants and longer follow-up time may better reflect the relationship between the use of GLP-1RAs and the risk of respiratory disease. It is meaningful to further evaluate the relationship between long-term use of GLP-1RAs and the risk of respiratory diseases in the future. In the secondary outcome, the risk of pulmonary edema and bronchitis was reduced. On the other hand, although statistical significance was not reached, the use of GLP-1RAs showed a trend towards a reduced risk of nine respiratory diseases (pneumonia, pulmonary fibrosis, dyspnea, acute respiratory failure, pleural effusion, COPD, sleep apnea syndrome, and pulmonary embolism).

GLP-1RAs were highly expressed in the lung [4]. Among the extra-pancreatic organs, the lung had the highest level of Glp1r mRNA expression [44, 45]. Moreover, alveolar lining fluid had GLP-1 concentrations that were substantially higher than those in serum [46]. This provides a pathway for a potential effect of GLP1-RA beyond glucose regulation. So far, the mechanisms underlying the protective effects of GLP-1RA in respiratory disease are not fully understood. GLP-1 RA may reduce tracheal hyperresponsiveness by activating the cAMP-dependent protein kinase A pathway in the human airways [4]. It also mediates tracheal relaxation. In a study using female mice in an obstructive pulmonary disease model, the use of GLP-1RA was found to reduce inflammation and improve lung function [10]. Meanwhile, some studies on GLP-1RA's anti-inflammatory effects in T2D patients have been conducted. GLP-1RA inhibits oxidative stress and inflammatory mediator production in patients with type 2 diabetes [47]. GLP-1RA dramatically lowered the levels of eosinophil surface activation markers after LPS stimulation. Thus, it attenuated LPS-stimulated eosinophil activation and reduced the generation of IL-4, IL-8, and IL-13 [12]. This suggests that GLP-1 may have an immunomodulatory role in asthma. In addition, in a previous animal model, liraglutide pretreatment attenuated LPS-induced acute lung injury in mice. It was linked to a decrease in the activity of inflammatory cytokines and chemokine genes in the lung, and it significantly decreased lung injury scores, and lung apoptosis [11]. In a related study, liraglutide was shown to increase FVC, and carbon monoxide diffusion capacity, and significantly reduce serum surface active protein D (SP-D) levels (used as a biomarker to assess alveolar-capillary barrier integrity) in patients with type 2 diabetes [48, 49]. It is well known that pulmonary surface-active substances (PS) are composed of various lipids and proteins arranged on the alveolar surface. Its main function is to maintain lung compliance and lung fluid homeostasis. on the other hand, surfactant protein A is the most prevalent protein in PS. As well, it has been shown that liraglutide increases the expression of SP-A by activating the TTF-1 signaling pathway. Thus, it may reduce the occurrence of pulmonary edema [50].

In the results of a cohort study published in BMJ, new users of GLP-1 receptor agonists had a 30% lower risk of severe COPD exacerbation compared to new sulfonylurea users. There was also a reduced risk of moderate exacerbation [6]. In another retrospective study, it was shown that patients starting GLP-1RA in intensive glucose-lowering therapy for patients with combined asthma and type 2 diabetes had fewer asthma attacks than those using other alternatives. Asthma symptoms were also reduced [7]. Interestingly, insulin resistance may enhance the pro-asthmatic effects of obesity. GLP-1 RAs may reduce asthma through weight loss and improved insulin sensitivity [51,52,53]. In addition, in a study that included 16,690 patients, patients with GLP-1RA experienced less worsening of chronic lower respiratory disease compared to DPP-4I users [8]. Finally, GLP-1 receptor agonists may improve lung function by reducing body weight, as abdominal obesity decreases lung capacity by pushing the diaphragm into the chest cavity.

In our analysis, a trend towards a decreased risk of sleep apnea syndrome was found with the use of GLP-1RA (RR 0.86, 95% CI 0.50–1.45). GLP-1 RAs may benefit patients with this obstructive sleep apnea syndrome by weight loss, reducing systemic inflammation, improving metabolic dysregulation, and reversing endothelial dysfunction [9]. Glucagon-like peptide-1 agonists have been advocated in the literature as a one-stop shop for the obesity-diabetes-OSA triad [9]. Besides, our results suggest that GLP-1RAs may not affect Pulmonary hypertension. However, in previous studies, it was demonstrated that GLP-1RAs could diastole pulmonary arteries in a cAMP-dependent manner [54]. Meanwhile, Glucagon-like peptide 1 receptor agonists attenuate autophagy via Drp1/NOX- and Atg-5/Atg-7/Beclin-1/LC3β pathways to improve pulmonary hypertension [55]. The inconsistent results may be due to the low incidence and the incorporation of studies with limited sample sizes, statistical effectiveness was not greatly increased. And it may lead to small sample effect bias. In the study of several categories of GLP-1 receptor agonists, we discovered that Liraglutide, Semaglutid, and Dulaglutide were connected to the low incidence of respiratory disease. This may be related to the high amino acid homology of natural glucagon-like peptide-1, with Liraglutide reaching 97% [56]. In addition, the chemical structures of different GLP-1 RAs with different pharmacological characteristics from each other. And these different findings could also be explained by unbalanced sample sizes.

This article has several advantages. Firstly, this is the first meta-analysis to evaluate the relationship between GLP-1RAs and the risk of respiratory disease in greater detail. The overweight or obese population was also included and subgroup analyses were carried out. In addition, all included studies were randomized controlled trials. Finally, there was either no heterogeneity or very little heterogeneity in our meta-analysis. We admit several limitations of our research. Almost all included studies did not have respiratory diseases as a primary outcome, and these outcomes were derived from adverse event reports while not detecting changes in lung function. Second, the analysis included only trials reporting respiratory disease, resulting in an unknown publication bias. In addition, although 28 studies were incorporated into this analysis, the wide confidence intervals reduced the quality of our findings. Finally, the limited amount of events in the subgroups may have led to a lack of certainty in the subgroup analysis. The effect of using GLP-1RAs on respiratory disease is unknown. More large long-term RCTs with respiratory disease as the primary outcome are needed in the future to evaluate the link between the incidence of respiratory disease and GLP-1 receptor agonists.

Conclusion

In conclusion, using GLP-1RAs was linked to a lower risk of overall respiratory diseases, especially Pulmonary edema and Bronchitis. In the future, physicians should pay attention to the relationship between GLP-1 RA use and the risk of respiratory diseases and evaluate the efficacy of GLP-1RAs in the primary and secondary prevention of respiratory diseases.