Mortality and other adverse outcomes in patients with type 2 diabetes mellitus admitted for COVID-19 in association with glucose-lowering drugs: a nationwide cohort study

The novel coronavirus disease 2019 (COVID-19) caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) initially emerged in China in December 2019 and quickly spread around the world, causing a global pandemic [1].

For patients with COVID-19, diabetes has been reported as one of the most frequent comorbidities, occurring in around 20% of patients [2, 3]. It is well known that type 2 diabetes mellitus (T2DM) may negatively impact clinical outcomes in patients with COVID-19. Possible poor outcomes include moderate and severe cases of COVID-19 disease, a higher rate of hospitalized patients in intensive care unit (ICU), a higher rate of treatment with anti-interleukin 6 receptor antibody (tocilizumab), and higher mortality [4]. Furthermore, it has been reported that hyperglycemia during COVID-19 infection, and particularly at hospital admission, has been associated with worse COVID-19 outcomes and could be a prognostic factor for worse outcomes in patients with or without T2DM [5, 6]. Indeed, patients with hyperglycemia may experience reduced effect of anti-COVID-19 therapies, particularly tocilizumab, an anti-interleukin 6 receptor antibody indicated for patients with moderate-to-severe COVID-19 pneumonia [7]. Thus, not only T2DM status but also the presence of hyperglycemia could have unfavorable effects on hospital admission, clinical outcomes, and drug therapy, leading to a worse prognosis in COVID-19 patients [5, 6].

However, to date, there is no conclusive evidence in regard to the potential implications on adverse outcomes of glucose-lowering drugs in patients with COVID-19 [8‐11]. In a recent study, the use of oral glucose-lowering medications (metformin, α-glucosidase, secretagogues, and dipeptidyl peptidase-4 inhibitors (DPP-4i)) showed neutral effects on in-hospital mortality and the composite outcome of poor prognosis defined as progression to severe or critical illness and in-hospital death. In contrast, insulin usage was associated with a greater risk of poor prognosis compared to not using it [12]. Other studies have reported a beneficial effect of metformin and DPP-4i on clinical outcomes in patients hospitalized with COVID-19 [13, 14]. The potential effects of these medications on COVID-19-related pathological mechanisms, such as systemic inflammation and endothelial dysfunction, have been proposed as the underlying mechanisms of possible benefits in reports on mechanistic hypotheses and preliminary data [10, 15, 16].

We conducted this study based on the premise that patients with T2DM seem particularly prone to a worse prognosis if infected by SARS-CoV-2, and in response to the fact that there is only controversial, limited evidence on the role of glucose-lowering drugs on clinical adverse outcomes. Our main objective was to examine the association between in-hospital death and each routine at-home glucose-lowering drug both individually and in combination with metformin in patients with T2DM admitted for COVID-19. We also evaluated their association with the composite outcome of need for ICU admission, invasive and non-invasive mechanical ventilation, or in-hospital death as well as on the development of in-hospital complications and a long-time hospital stay.

Our study found that none of the at-home glucose-lowering drugs analyzed (metformin, DPP-4i, insulin, metformin plus DPP-4i, metformin plus SGLT-2i, and metformin plus insulin) showed a significant association with in-hospital deaths; the composite outcome of the need for ICU, mechanical ventilation, or in-hospital death; in-hospital complications; or a long-time hospital stay.

Currently, only limited evidence is available on the role of glucose-lowering drugs in adverse clinical outcomes. Recently, a study conducted by Chen et al. [12] focused on the impact of glucose-lowering medications on the clinical outcomes of patients with diabetes and COVID-19, showing that insulin users had a greater risk of a composite outcome of progression to severe or critical illness and in-hospital death compared with non-insulin users (OR 3.58, 95% CI 1.37–9.35, p = 0.009). However, none of the glucose-lowering medications (metformin, DPP-4i, α-glucosidase, secretagogues, and insulin) was associated with in-hospital deaths. In our study, all glucose-lowering medications evaluated, including oral drugs and insulin, had no significant association with in-hospital adverse outcomes in patients admitted for COVID-19. The differences in the definitions of the outcomes of clinical complications during the hospitalization, the methodology used, and the sample size could explain the difference found regarding the role of insulin. We included an important number of patients taking glucose-lowering drugs and performed a propensity score matching that included variables that could affect the treatment choice or adverse outcomes as independent variables.

Two observational studies have specifically evaluated the role of metformin and DPP-4i in patients admitted for COVID-19 [13, 14]. The study focused on metformin, conducted by Bramante et al. [13] in the USA, found no significant reduction in in-hospital mortality in the overall sample, which is in line with our results. However, when a subgroup analysis was performed, women with obesity and T2DM who used metformin were observed to have a significantly lower mortality rate. Metformin has been proposed to reduce levels of IL-10, IL-6, and TNFα—important mediators in macrophage activation and cytokine release; improves neutrophil/lymphocyte ratio; stabilizes mast cells; decreases thrombosis; and improves endothelial function, thus reducing the adverse impacts of mortality and complication in patients with COVID-19 [13, 22‐27]. In the study of Rhee et al. [14] from Korea, protective effects against severe/lethal cases were shown in patients with T2DM who used DPP-4i in monotherapy or in combination with renin-angiotensin system blockers after adjusting for age, gender, comorbidity, and medications [14]. It is known that DPP-4i have anti-fibrotic activity and modulate inflammation and that these properties could be helpful in reducing the progression towards a hyperinflammatory state associated with severe COVID-19 [14, 28, 29]. All these major inflammatory mechanisms observed in some COVID-19 patients suggest that the endothelium could be a key target organ in COVID-19 infection [30]. The role of renin-angiotensin system blockers in hypertensive patients with COVID-19 has also been widely discussed [31]. Although hypertension has been associated with a higher risk of mortality, recent studies have shown that previous treatment with angiotensin-converting enzyme inhibitors and angiotensin II receptor blockers did not alter the outcomes in hypertensive patients [31, 32]. In our study, no significant associations with adverse clinical outcomes were found among users of DPP-4i and users of the other glucose-lowering drugs. This difference could be explained by the different methodology used in our study, which includes an important adjustment using a propensity score matching in order to compare the two groups of glucose-lowering drugs, as well as the differences in the baseline characteristics, with older patients and higher comorbidity found in our study compared with Rhee et al.’s study (mean age of 74.9 ± 8.4 vs 63.7 ± 12.2 years old).

Other glucose-lowering treatments in our study, such as metformin in combination with SGLT-2i or with insulin, showed no significant differences compared with the other glucose-lowering treatments. No propensity score matching was performed for these two comparisons due to the small number of patients in these treatment groups, so no solid conclusions may be drawn. To our knowledge, no evidence exists about the role of SGLT-2i on adverse outcomes in patients with COVID-19. A hypothetical anti-viral effect of SGLT-2i has been suggested, as these agents can increase lactate concentrations and decrease intracellular pH, which could reduce the viral load [33].

Our findings are important because they provide valuable information on the role of at-home glucose-lowering drugs on adverse outcomes in patients with T2DM admitted for COVID-19. In addition, this is the first study to report the role of each glucose-lowering drugs alone or in combination with metformin compared to the other glucose-lowering drugs after a robust adjustment using a wide number of confounding variables. Furthermore, data were collected in a large multicenter, nationwide study. Nevertheless, these results should be considered within the context of several potential limitations. Despite the propensity matching analysis performed and due to the fact that the data were obtained retrospectively via medical records from the electronic medical record system, the possible effects of unmeasured confounding factors cannot be excluded. In addition, some glucose-lowering drug group comparisons had a small number of patients after propensity score matching that could be an explanation for the lack of significant differences. No other intermediate adverse outcomes were explored due to their reduced number. Furthermore, we did not record the characteristics of T2DM, such as glycemic control before hospitalization, duration of diabetes, blood glucose levels during the hospitalization, or in-hospital anti-hyperglycemic management. Lastly, the data provided about at-home glucose-lowering drugs did not include information on treatment adherence or treatment duration.

Our study found that the at-home use of metformin, DPP-4i, insulin, metformin plus DPP-4i, metformin plus SGLT-2i, and metformin plus insulin showed no significant association with in-hospital deaths; the composite outcome of need for ICU admission, mechanical ventilation, or in-hospital death; in-hospital complications; or long-time hospital stay in patients with T2DM admitted for COVID-19. Given the large number of patients with T2DM and COVID-19 infection, the ominous relationship between these pathologies, and the limited evidence on the role of glucose-lowering drugs, prospective studies are needed.