Our meta-analysis aimed to review the utility of SGLT-2 inhibitors in COVID-19 patients comprehensively. Focusing on their role as an intervention for hospitalised COVID-19 patients, our analysis centered on two randomized controlled trials: the RECOVERY and DARE-19 trials31,32. The analysis revealed no significant differences between the two groups in terms of mortality, hospital discharge, need for mechanical ventilation, renal replacement therapy, acute kidney injury, or ketoacidosis associated with SGLT2 inhibitor use. Both the RECOVERY and DARE-19 trials reached similar conclusions, although their inclusion criteria varied31,32. The RECOVERY trial did not include patients who had diabetes at the time of recruitment but did include patients with a history of diabetes31. In contrast, the DARE-19 trial included participants exhibiting cardiometabolic risk factors, including type-2 diabetes, heart failure, and chronic kidney disease32. The presence of comorbidities, particularly diabetes, heart failure, and chronic kidney disease, may have influenced the impact of the drug on disease progression, considering the established efficacy of SGLT2 inhibitors in these conditions2,50. However, in the RECOVERY trial, a post-hoc subgroup analysis was performed between patients with and without diabetes, yielding insignificant results. Despite the given evidence, trials are scarce to derive a definitive conclusion. Our analysis of hospital discharge within one month yielded a pooled risk ratio of 1.02 (CI: 1.00; 1.05), with the lower limit of the confidence interval intersecting 1, underscoring the necessity for further trials to establish a conclusive inference.
We performed a meta-analysis of eligible observational studies to further understand the effects of SGLT-2 inhibitors in COVID-19 patients with a history of SGLT-inhibitor use. Among the 17 observational studies analysed, 12 focused on mortality outcomes. These 12 studies collectively included 20,418 COVID-19 patients with prior SGLT2i use and 104,254 COVID-19 patients without prior SGLT2i exposure. The analysis revealed a significant 33% reduction in mortality odds among individuals with prior SGLT2i use compared to non-SGLT2i users. Except for the study by Kahkoska et al., all studies in the mortality analysis specifically included populations with pre-existing diabetes, potentially introducing bias with regard to our research question. This suggests that the effects of medication could be attributed to its impact on diabetes rather than solely on COVID-19. However, drawing a definitive conclusion regarding this association remains challenging because of an inherent limitation of the available data.
Similarly, studies by Khunti et al., Shestakova et al., and Solerte et al. reported favorable outcomes in COVID-19 patients with T2DM on SGLT2 inhibitors. Khunti et al., ‘s retrospective analysis involving nearly 3 million individuals showed a reduced COVID-19-related mortality risk (adjusted hazard ratio: 0.82) for SGLT2i users compared to those not on glucose-lowering medications39,45,51. The study of Shestakova et al. highlighted a significantly lower fatality rate among T2DM patients on SGLT2 inhibitors before infection. Solerte et al.’s’ recent research indicated lower death rates in COVID-19 patients with diabetes receiving SGLT2i therapy upon hospital admission51.
A comparison of these medications provides insights into their relative efficacy.. Shestakova et al.'s study showed varied mortality outcomes between different glucose-lowering therapies among type 2 diabetic patients. Specifically, glucagon-like peptide-1 receptor agonists exhibited a lower case fatality rate (CFR) than SGLT2 inhibitors45. In contrast, Israelsen et al. reported comparable 30-day mortality rates between users of GLP-1 receptor agonists (GLP-1 RA) and SGLT-2 inhibitors (3.3% vs. 3.7%), and both were lower than those for DPP-4i users (8.6%) 34. Khunti et al., however, concluded that there was little overall variation in outcomes for all glucose-lowering treatments, including SGLT2is, and that these differences were probably confounded by indication39. Our subgroup analysis confirmed that SGLT2is have a better outcome than GLP-1 RA or DPP-4i, which aligns with all observational studies analysed.
The assessment of hospitalisation risk in COVID-19 patients across observational studies revealed a 16% reduction in odds. When SGLT2 inhibitors were directly compared with GLP-1 RA/DDP-i, there was a 17% decrease in odds. Israelsen et al., however, reported that hospital admission risks were generally comparable for GLP-1 RA and DPP-4i users and SGLT-2i users 34. However, their study had a smaller sample size than those analysed for this outcome. Meta-analysis of ICU admission and mechanical ventilation across studies yielded statistically insignificant results, indicating no significant difference in odds ratios between the various treatments. These findings suggest that, while SGLTi offers a potential advantage in reducing the risk of hospitalisation, it does not seem to confer additional benefits concerning ICU admission. It is plausible that patients requiring ICU care might already be in a more critical health state, where the improvement with these oral hypoglycaemic agents becomes less discernible or less effective. This implies that medication might be more influential in mitigating less severe COVID-19 cases.
Understanding the potential efficacy of SGLT2i in addressing COVID-19 infection and its complications is crucial. Despite their primary use in diabetes management, drugs, such as dapagliflozin, show additional beneficial properties beyond glucose control. These medications have pleiotropic effects, including anti-inflammatory actions, enhanced myocardial function, and improved oxygen delivery, as documented in previous studies 52,53. SGLT2i's role in COVID-19 can be understood through multiple mechanisms (Figure 4a). One of them that emerges prominently is its influence on cellular pH levels. The binding of SARS-CoV-2 to the angiotensin-converting enzyme 2 (ACE2) is facilitated in acidic environments54. SGLT2i indirectly regulates this pH balance, potentially reducing the viral load. Additionally, SGLT2i upregulates ACE2 receptors, increasing Ang 1-7 levels, which are protective against acute respiratory distress syndrome (ARDS) caused by COVID-1955–57.
SGLT2 inhibitors also significantly reduced inflammation and oxidative stress, which are two key factors in COVID-19's pathogenesis58. By lowering proinflammatory cytokines such as IL-6 and TNF-a, SGLT2i could mitigate the cytokine storm associated with severe COVID-19 cases59. Furthermore, these inhibitors impact myocardial and endothelial function, reducing the likelihood of complications such as myocarditis and heart failure60,61. SGLT2 inhibitors also induce metabolic alterations that influence the disease. For instance, they decrease lactate production and modulate the activity of the lactate/H+-symporter and Na+/H+ exchanger (NHE), reducing cellular swelling and death62,63. This mechanism is important because SARS-CoV-2 infection often leads to tissue hypoxia and heightened lactate production 64. Of note, dapagliflozin has been shown to lower the incidence of new-onset diabetes in COVID-19 patients65. This finding holds significance as new-onset diabetes is a recognised complication of COVID-19, possibly attributed to viral-induced damage to pancreatic cells 66,67. Moreover, SGLT2i might have a protective role against COVID-19-related renal complications. Acute kidney injury (AKI) is a common complication in COVID-19 patients68,69, and SGLT2i's nephroprotective properties could prove beneficial in these cases69–72. Figure 4b shows an illustrative diagram depicting the potential mechanisms through which SGLT2 inhibitors can benefit COVID-19 patients.
However, it is crucial to consider the balance between these advantages and potential risks associated with SGLT2 inhibitors. SGLT2i can increase the likelihood of diabetic ketoacidosis (DKA), especially regarding COVID-19 infection, which can trigger DKA73. Additionally, they may exacerbate low blood pressure in severe COVID-19 cases and increase the risk of urogenital fungal infections, particularly when combined with treatments such as dexamethasone and antibiotics commonly used in COVID-1974–78. However, our analysis of diabetic ketoacidosis did not show a statistically significant association, suggesting that the observed risk of DKA with SGLT2 inhibitors in the context of COVID-19 requires further investigation.
The Umbrella review part of our study comprised 10 reviews: one systematic review of case reports79, six systematic reviews and meta-analyses7,8,80–83, one meta-analysis84, and two network meta-analyses4,57. The review population consisted of patients with T2DM who contracted COVID-19 and were on SGLT-2 inhibitors either before or after their hospitalisation or the onset of infection. The reviews primarily assessed outcomes such as the severity of COVID-19 infection7,80,82, need for Intensive Care Unit (ICU) admission, invasive and non-invasive mechanical ventilation4,7,8,80,82, Disseminated Intravascular Coagulation (DIC)82, Euglycemic Diabetic Ketoacidosis (Eu-DKA)79,81,83 and in-hospital Mortality4,7,8,57,80–84.
The reviews suggest that prior use of SGLT-2i was associated with reduced adverse outcomes related to COVID-19 and notably lowered mortality rates among individuals with diabetes.7,8,57,81,82,84. It was also noted that there was a decreased risk of hospitalisation54 and a reduced risk of requiring non-invasive or invasive mechanical ventilation61. While a review of case reports suggested a possible increase in the risk of euglycemic diabetic ketoacidosis79 with SGLT2i, the level of evidence was deemed non-significant. A systematic review and meta-analysis supported the notion that SGLT-2 inhibitors reduce the mortality in COVID-19 without increasing the risk of DKA81. Conversely, another review found that prior SGLT-2 inhibitor treatment, when associated with euglycemic Diabetic Ketoacidosis, might protect renal function in COVID-19 patients with pre-existing ketotic states.83. The reviews mostly fell into Class III 4,8,81,83 and Class IV7,57,80,82,84 levels of evidence, with one review deemed non-significant79. Our analysis of both observational studies and trials showed no significant increase in the odds of developing DKA. Based on this collective assessment, it appears that SGLT2 inhibitors may not inherently lead to DKA.
Although SGLT2 inhibitors such as dapagliflozin offer promising therapeutic avenues to manage COVID-19 and its complications, carefully considering their risks and benefits is essential. Their role in modulating pH levels, reducing inflammation and oxidative stress, and protecting against metabolic, cardiac, and renal complications positions them as valuable pharmacological agents for COVID-19 management. However, their utilization requires careful consideration, particularly in severely ill COVID-19 patients.