key: cord-0773056-ow0q72f5
authors: Laurenzi, Andrea; Caretto, Amelia; Molinari, Chiara; Mercalli, Alessia; Melzi, Raffaella; Nano, Rita; Tresoldi, Cristina; Rovere Querini, Patrizia; Ciceri, Fabio; Lampasona, Vito; Bosi, Emanuele; Scavini, Marina; Piemonti, Lorenzo
title: No Evidence of Long-Term Disruption of Glycometabolic Control After SARS-CoV-2 Infection
date: 2021-10-29
journal: J Clin Endocrinol Metab
DOI: 10.1210/clinem/dgab792
sha: 89be32ac26edcd13e02de7a05dacf217cb0af800
doc_id: 773056
cord_uid: ow0q72f5

PURPOSE: To assess whether dysglycemia diagnosed during severe acute respiratory syndrome coronavirus 2 pneumonia may become a potential public health problem after resolution of the infection. In an adult cohort with suspected coronavirus disease 2019 (COVID-19) pneumonia, we integrated glucose data upon hospital admission with fasting blood glucose (FBG) in the year prior to COVID-19 and during postdischarge follow-up. METHODS: From February 25 to May 15, 2020, 660 adults with suspected COVID-19 pneumonia were admitted to the San Raffaele Hospital (Milan, Italy). Through structured interviews/ medical record reviews, we collected demographics, clinical features, and laboratory tests upon admission and additional data during hospitalization or after discharge and in the previous year. Upon admission, we classified participants according to American Diabetes Association criteria as having (1) preexisting diabetes, (2) newly diagnosed diabetes, (3) hyperglycemia not in the diabetes range, or (4) normoglycemia. FBG prior to admission and during follow-up were classified as normal or impaired fasting glucose and fasting glucose in the diabetes range. RESULTS: In patients with confirmed COVID (n = 589), the proportion with preexisting or newly diagnosed diabetes, hyperglycemia not in the diabetes range and normoglycemia was 19.6%, 6.7%, 43.7%, and 30.0%, respectively. Patients with dysglycemia associated to COVID-19 had increased markers of inflammation and organs’ injury and poorer clinical outcome compared to those with normoglycemia. After the infection resolved, the prevalence of dysglycemia reverted to preadmission frequency. CONCLUSIONS: COVID-19–associated dysglycemia is unlikely to become a lasting public health problem. Alarmist claims on the diabetes risk after COVID-19 pneumonia should be interpreted with caution.

FBG prior to admission and during follow-up were classified as normal or impaired fasting glucose and fasting glucose in the diabetes range. Results: In patients with confirmed COVID (n = 589), the proportion with preexisting or newly diagnosed diabetes, hyperglycemia not in the diabetes range and normoglycemia was 19.6%, 6.7%, 43.7%, and 30.0%, respectively. Patients with dysglycemia associated to COVID-19 had increased markers of inflammation and organs' injury and poorer clinical outcome compared to those with normoglycemia. After the infection resolved, the prevalence of dysglycemia reverted to preadmission frequency. Conclusions: COVID-19-associated dysglycemia is unlikely to become a lasting public health problem. Alarmist claims on the diabetes risk after COVID-19 pneumonia should be interpreted with caution.

Key Words: COVID-19, diabetes, humans Individuals with either type 1 or type 2 diabetes are more likely to develop severe coronavirus disease 2019 and to die from severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection than people who do not have diabetes (1) (2) (3) (4) (5) (6) (7) . Moreover, acute metabolic decompensation of preexisting diabetes is a well-recognized complication of SARS-CoV-2 infection (8, 9) . Furthermore, isolated cases of new-onset diabetes have been reported during COVID-19 (10) (11) (12) (13) (14) (15) (16) , suggesting the possibility that SARS-CoV-2 might exert direct cytotoxicity against beta cells with a diabetogenic effect (17, 18) . Intriguingly, in vitro studies reported that the putative receptor angiotensin converting enzyme-2 (ACE2) and the effector protease transmembrane serine protease 2 (TMPRSS2), 2 factors associated with SARS-CoV-2 infection (19) , are expressed in isolated human islets and in pancreatic endocrine cells derived from human pluripotent stem cells (20, 21) . This would be consistent with the observation that, using both pseudoviral particles and live SARS-CoV-2, pancreatic alpha and beta cells appear permissive to SARS-CoV-2 infection (21, 22) . Finally, the existence of a SARS-CoV-2-induced beta-cell transdifferentiation was recently suggested (23) . However, other studies have disproved a significant expression of ACE2 and TMPRSS2 in native human pancreatic islet beta cells from normal donors, suggesting that a direct diabetogenic effect of SARS-CoV-2 via ACE2 and TMPRSS2 is unlikely (24, 25) . Despite these discrepancies, all studies agreed that SARS-CoV-2 can be detected in the pancreas where it may cause inflammation and indirectly affect beta cells. Consistently, pancreatic enlargement, abnormal amylase or lipase levels, and pancreatitis were described in critically ill COVID-19 patients (26) (27) (28) . However, the evidence that different pancreatic tissues are susceptible to SARS-CoV-2 infection does not automatically imply that SARS-CoV-2 infection will have a permanent effect on glucose homeostasis after infection resolution or will trigger a permanent diabetes.

Recently, Montefusco et al (29) reported that among patients with new-onset hyperglycemia at hospital admission for COVID-19, persistent hyperglycemia continued to be observed in the 6 months after resolution of SARS-CoV-2 infection in 20 out 57 of patients (35.1%), and overt diabetes was diagnosed in 1 out of 57 of patients (1.6%), suggesting the persistence of aberrant glycometabolic control long after recovery from COVID-19. Using a classification of dysglycemia associated with SARS-CoV-2 infection that differs from that of Montefusco et al (29) , we previously reported in a cohort of patients with confirmed COVID-19 that only a small percentage of patients without preexisting diabetes maintained or achieved aberrant glycometabolic control during follow-up, possibly unrelated to the SARS-CoV-2 infection (30) . To address the discrepancy between these 2 studies, we analyzed glucose data at the time of admission in a cohort of 660 adult cases with suspected COVID-19. We classified their dysglycemia according to the criteria used by Montefusco et al (29) and assessed whether this persists throughout follow-up or reverts when the viral infection resolves.

The study population consisted of 660 adults (aged ≥18 years) with suspected COVID-19 pneumonia admitted between February 25 and May 15, 2020, to the emergency or clinical departments of the IRCCS San Raffaele Hospital (Milan, Italy) and for whom a serum sample was stored in our institutional biobank (31) . This series of patients is part of an institutional clinical-biological cohort (COVID-BioB; ClinicalTrials.gov Identifier: NCT04318366) of patients with COVID-19 (32) . The Institutional Review Board (protocol number 34/int/2020) approved the study. Informed consent was obtained by all participants according to Institutional Review Board guidelines.

A confirmed COVID-19 case was defined as previously described (33) . Briefly, a patient with SARS-CoV-2 positive reverse-transcriptase polymerase chain reaction (rtPCR) from a nasal/throat swab and signs, symptoms, and radiological findings suggestive of COVID-19 pneumonia was classified as positive (n = 558). In case of multiple (at least 2) SARS-CoV-2 negative rtPCR in the presence of radiological findings suggestive of COVID-19 pneumonia, subjects were classified as having a confirmed infection if they were positive for immunoglobulin M/immunoglobulin G against SARS-CoV-2 spike protein (34) (n = 28). SARS-CoV-2 infection was excluded in subjects with multiple (at least 2) SARS-CoV-2 negative rtPCR and negative for immunoglobulin M/immunoglobulin G against SARS-CoV-2 spike protein (n = 74). Demographic information, clinical features, and laboratory tests were obtained within 72 h from admission. Data were collected as previously described (33) during hospitalization or after discharge from both structured baseline patient interviews and hospital paper or electronic medical records. We also reviewed the electronic medical records in the year prior to the hospital admission for COVID-19.

Routine blood tests encompassed serum biochemistry [including renal function and lactate dehydrogenase (LDH)], complete blood counts with differential, inflammation markers [C-reactive protein (CRP), ferritin, interleukin-6 (IL-6)], and D-dimer.

Fasting blood glucose (FBG) were measured in each patient at different time points after admission to the emergency room and during hospitalization. A mean of 4 ± 1.6 (SE) time points per patients were available. In the year prior to hospital admission, we retrieved FBG for 234 out of the 660 patients in our cohort. Glycated hemoglobin (HbA1c) levels during in hospital stay or in the year prior to hospital admission were also abstracted, where available. We also recorded FBG, body mass index (BMI), and HbA1c during the postdischarge follow-up (ie, at the 1-, 3-, 6-, and 12-month outpatient visits), where available. Regarding glucose treatment modalities applied during the hospitalization period, as no specific guidelines were available during the first wave, patients were treated as suggested by American Diabetes Association (ADA) Standards of Medical Care in Diabetes-2020 (35): insulin therapy was initiated for treatment of persistent hyperglycemia starting at a threshold ≥180 mg/dL (10.0 mmol/L). Once insulin therapy was started, a target glucose range of 140 to 180 mg/dL (7.8-10.0 mmol/L) was recommended.

We stratified glucose abnormalities at the time of hospitalization using the criteria recently reported by Montefusco et al (29) . Briefly, study participants were defined as having (1) preexisting diabetes if prior to hospital admission for COVID-19 they had a documented diagnosis of diabetes or were prescribed diabetes medications; (2) newly diagnosed diabetes if they had a negative history of diabetes, no prescription of diabetes medications, and a FBG during hospitalization, in the absence of infusions of dextrose, of 7.0 mmol/L or higher (ADA criteria); (3) hyperglycemia not in the diabetes range if they had random blood glucose levels between 100 and 199 mg/dL or 2 FBG >100 mg/dL and <126 mg/dL; or (4) normoglycemia if they had no history of diabetes and had normal glucose levels according to the ADA criteria. FBG recorded before admission and during the follow-up were classified as normal fasting glucose, impaired fasting glucose (IFG), or fasting glucose in the diabetes range (DFG) according to ADA criteria.

Categorical variables are reported as frequency or percentage, continuous variables as median with interquartile range, or mean with SD. Categorical variables were compared using Chi-square or Fisher's exact test, as appropriate. Continuous variables were compared using paired/ unpaired t-test, analysis of variance (Bonferroni post hoc test), Mann-Whitney test, Kruskal-Wallis test (Dunn's post hoc test), and Wilcoxon, as appropriate. Spearman correlation was performed to assess the association between glycemia and biochemistry variables. The time to event was estimated according to Kaplan-Meier, from the date of symptom onset to the date of the event or of last follow-up visit, whichever occurred first. We used univariate and multivariate Cox proportional hazards models to study the association between patient characteristics with time to adverse outcome (a composite endpoint of admission to intensive care unit or death, whichever occurred first). The effect estimates were reported as hazard ratios (HRs) with the corresponding 95% CI estimated according to the Wald approximation. Two-tailed P-values are reported, with P-value < 0.05 indicating statistical significance. All CIs are 2-sided and not adjusted for multiple testing unless otherwise specified. Statistical analyses were performed with SPSS 24 (SPSS Inc./IBM) and GraphPad Prism version 5.04.

From February 25 to May 15, 2020, 660 adults with suspected COVID-19 pneumonia were enrolled in our institutional clinical-biological cohort (COVID-BioB).

A diagnosis of COVID-19 was confirmed in 586 out of 660 cases (88.8%, COVID cohort), while the SARS-CoV-2 infection was excluded in the remaining 74 cases (5.1%, No-COVID cohort). In the COVID cohort, 154 out of 586 patients (26.3%) presented with increased blood glucose levels (Fig. 1A) . One hundred and fifteen out of 154 (19.6%) had preexisting diabetes, while a new diagnosis of diabetes was made in the remaining 39 patients (6.7%) during their in hospital stay. Hyperglycemia not in the diabetes range was observed during hospitalization in 256 out of 586 subjects (43.7%), without preexisting or newly diagnosed diabetes, while the remaining 176 patients (30%) had normal blood glucose levels throughout their hospital stay. Among the 74 patients in whom COVID-19 was not confirmed, 17 (23.0%) had diabetes, either preexisting diabetes (n = 10, 13.5%) or newly diagnosed diabetes (n = 7, 9.5%); 15 (20.2%) had hyperglycemia not in the diabetes range; and 42 (56.8%) were normoglycemics throughout their hospital stay (Fig. 1A) . Demographic and clinical characteristics of our study population are reported in Table 1 , where we combined patients with preexisting diabetes and newly diagnosed diabetes. As expected, mean HbA1c levels were significantly higher in patients with diabetes compared to patients with hyperglycemia not in the diabetes range or normoglycemia but not different between patients with hyperglycemia and normoglycemia (Fig. 1B) . On the other hand, mean FGB and peak blood glucose levels during hospital stay were different between groups, with the highest levels in patients with diabetes (Fig. 1C) . Upon admission patients with diabetes exhibited significantly higher white blood cell count [7.9 (5.8-11.7) vs 6.4 (4.7-9) × 10 9 /L, P < 0.0001], neutrophil-to-lymphocyte ratio (Fig. 2) . Generally, patients with hyperglycemia not in the diabetes range showed intermediate levels between patients with diabetes and patients with normoglycemia (Fig. 2) . IL-6 levels were not different between groups. In the overall population, Spearman's correlation showed negligible or weak correlation, although statistically significant, between mean FBG or peak glycemia values during hospitalization and the biochemistry variables analyzed [ Supplementary Figure 1 (36) ]. Glucose abnormalities were strongly associated with an increased risk of adverse clinical outcome, as defined by composite endpoint of admission to intensive care unit or death, whichever occurred first (Fig. 1D-1E , Table 1 ). Sex-and age-adjusted Cox proportional hazards model showed an increased risk for the adverse clinical outcome in patients with diabetes compared to those with normoglycemia [HR 3.28 (95% CI 2.04-5.3); P < 0.001] or hyperglycemia not in the diabetes range [HR 1.59, (95% CI 1.14-2.2); P = 0.006]. Moreover, an increased risk of adverse clinical outcome was evident also for patients with hyperglycemia not in the diabetes range compared to normoglycemics ones [HR 2.16 (95% CI 1.35-3.4); P = 0.001]. A multivariate analysis, including BMI, creatinine hypertension, cardiovascular disease, and known markers of diseases severity at admission (LDH, CRP, white blood cell), confirmed diabetes as an independent predictor of adverse clinical outcome (Fig. 1E) . Finally, patients with diabetes and hyperglycemia not in the diabetes range had a longer hospital stay compared to patients with normoglycemia (Fig. 1F) . In the COVID cohort, we analyzed FBG prior to COVID-19 hospitalization and after hospital discharge in the different glucose categories at the time of hospitalization for COVID-19 (Fig. 3) . Among 115 patients with preexisting diabetes, in the year prior to admission FBG was available for 50 subjects (43.5%) and during follow-up for 63 out of 76 surviving patients (82.1%). As expected, most patients showed DFG both before and after SARS-CoV-2 infection, without significant changes during the time. During the follow-up, no diabetic patients required drug treatment escalation. A marginal trend in increasing FBG was evident during hospitalization [from 137 (116-172) mg/dL to 150 (115-230) mg/dL, P = 0.11], followed by return to preinfection levels during follow-up [134 (108-169) mg/dL, P = 0.84]. Among 295 patients with newly diagnosed diabetes or hyperglycemia not in the diabetes range, FBG prior to COVID-19 was available for 84 patients (28.5%) and FBG during follow-up for 178 out of the 238 surviving patients (74.8%). Of note, in this group in 33.3% of patients IFG/DFG was already present prior to SARS-CoV-2 infection. A significant increase of FBG was evident during hospitalization [from 95 (85-107) to 102 mg/dL (93-115), P = 0.003] followed by return to preinfection levels during follow-up [97.5 (90-107) mg/dL; P = 0.24 vs preinfection levels, P = 0.004 vs hospitalization]. Among the 176 patients with normoglycemia during hospitalization, FBG prior to COVID-19 was available for 65 subjects (36.9%) and glucose measurements during follow-up for 112 out of the 161 surviving patients (69.6%). In this group a significant decrease of FBG was evident during hospitalization [from 91 (86-96.5) to 85 (78-89) mg/dL, P < 0.001] followed by a return to preinfection levels during follow-up [94 (88-100) mg/dL; P = 0.72 vs preinfection levels, P < 0.001 vs hospitalization]. A reduction in BMI from baseline was reported at the last observation in patients with newly diagnosed diabetes or hyperglycemia not in the diabetes range during hospitalization (Fig. 4) . On the contrary, a modest increase in BMI during follow-up was evident in patients with normoglycemia during hospitalization. HbA1c did not change significantly during the observation period in all 3 groups (Fig. 4) . The same analysis was performed also in the No-COVID cohort with similar results, even if the low number of patients in this cohort does not allow to draw definite conclusions [ Supplementary Figure 2 (36) ].

Whether dysglycemia associated with SARS-CoV-2 infection should be considered a specific clinical entity, and whether it persists or reverts when the viral infection resolves is still a matter of discussion. To address this issue, we studied a cohort of 621 adults patients with suspected COVID-19 pneumonia, assessing dysglycemia not only at the time of hospitalization but also in the year prior to COVID-19 hospitalization and during postdischarge follow-up, when FBG were available. Our study generated several interesting findings. (25) 28 (11) 19 (11) <0.001 First, the dysglycemia associated with COVID-19 (ie, newly diagnosed diabetes or hyperglycemia not in the diabetes range) resolved in the majority of patients after the SARS-CoV-2 infection resolved. A similar behavior was also evident in the No-COVID cohort, and it is, therefore, reasonable to speculate that reversible transient factors, such as inflammation-induced insulin resistance, may be causing dysglycemia in patients with pneumonia from SARS-CoV-2 or other pathogens (37) . This hypothesis is supported by the association between biochemical markers of inflammation during hospitalization and the degree of dysglycemia. Therefore, this finding does not support the persistence of a long-term disruption of glycometabolic control after SARS-CoV-2 infection, as claimed by other authors (29) . One possible explanation for this discrepancy is a potential assessment bias (ie, how patients with dysglycemia are identified and classified) in observational studies conducted under emergency conditions. Obviously, a biased stratification of the groups leads to biased conclusions, independently by the quality of data and the statistical analysis performed. A negative history of diabetes or prediabetes cannot rule out the presence of undiagnosed diabetes. Thanks to the access, although retrospective, to FBG values in the year prior to hospitalization for COVID-19, in our cohort we could document that DFG or IFG were often preexistent in patients that at hospital admission for suspected COVID-19 were classified as normoglycemics or with newly diagnosed diabetes or with hyperglycemia not in the diabetes range. Moreover, the prevalence of IFG, DFG, and FBG, as well as median HbA1c and BMI before and 6 months after SARS-CoV-2 infection (ie, outside an acute illness) were similar in the 3 groups.

Second, the prevalence of preexisting diabetes was similar in patients of the COVID and No-COVID cohorts, while the prevalence of newly diagnosed diabetes and hyperglycemia not in the diabetes range was higher in the COVID-19 cohort. This suggest that diabetes per se, as risk factor, is not unique to COVID-19. On the other hand, the well-known high level of inflammatory stress associated with the acute phase of COVID-19 pneumonia may account for the excess difference in newly diagnosed diabetes and hyperglycemia not in the diabetes range between the COVID and No-COVID cohort. Third, the prevalence of the dysglycemia associated with SARS-CoV-2 infection may vary significantly according to the criteria used to define glucose abnormalities. In this study, overt hyperglycemia was present in 43.7% of patients during hospitalization. This prevalence is comparable to that reported by Montefusco et al (29) (not unexpectedly, having used the same classification criteria) but certainly higher than that previously reported by us (38) or by other groups (39) using other classification. This underlines the need for standardizing the criteria to define the different categories of glucose abnormalities to allow comparison of the results from different studies.

Fourth, regardless of its reversibility, our study documented that dysglycemia was associated with an adverse clinical outcome of SARS-CoV-2 infection and that this association was independent from the major risk factors for disease severity. As good glycemic control was indeed shown to reduce disease severity and mortality in COVID-19 patients with hyperglycemia (40), early recognition and treatment of COVID-associated hyperglycemia may greatly benefit these patients. Therefore, our findings do strongly support the need to screen all patients with COVID-19 pneumonia for hyperglycemia (ie, with either fasting blood glucose and/or HbA1c) at the time of hospital admission, despite a mute personal or family history of diabetes.

Whether COVID-associated dysglycemia is a new clinical entity is still a matter of discussion. In terms of pathophysiology, dysglycemia may include (1) stress-induced hyperglycemia; (2) previously unrecognized (pre)diabetes (either type 2 or, less likely, type 1 diabetes); (3) a form of diabetes due to the direct effect of the SARS-CoV-2 on the pancreas; or (4) a form of diabetes secondary to the treatment of COVID-19 (eg, with glucocorticoids or other medications possibly inducing secondary hyperglycemia/ diabetes) (37, (41) (42) (43) (44) (45) (46) (47) . In our cohort, enrolled during the first wave of the pandemic, the use of glucocorticoids was very limited and, therefore, unlikely to have played a role in the pathogenesis of hyperglycemia. Since FBG reverted to preadmission values in most patients when the infection resolved, it is reasonable to hypothesize that reversible transient factors, such as inflammation-induced insulin resistance, may have played a major role in causing dysglycemia (37) . In any case, we cannot exclude that more than 1 cause may have contributed to the dysglycemia associated with SARS-CoV-2. Definitively addressing the specific role of SARS-CoV-2 in COVID-associated dysglycemia would require the systematic availability of pancreatic tissue and advanced studies on insulin secretion and resistance before, during, and after the infection, studies that would be extremely difficult or unrealistic to perform. Recently, in contrast to our data, some retrospective cohort studies have reported excess risk and relative hazards for developing incident diabetes after the acute phase of SARS-CoV-2 infection in adults (48, 49) . Significant risk differences for type 2 diabetes after SARS-CoV-2 infection ranged from 0.47 to 0.82 per 100 people, and HRs ranged from 1.39 to 1.83. The major limitation of these studies is that the incident new cases in the postacute phase may be just the identification of a preexisting condition that was simply undiagnosed before. In fact, epidemiological evidence in many countries suggests that for any 2 patients with diagnosed diabetes there might be 1 with undiagnosed diabetes. Moreover, those studies identified an excess risk for diabetes that was not unique to SARS-CoV-2, since it was also observed with other serious viral infections. For example, compared to individuals who were hospitalized with seasonal influenza, patients who were hospitalized for COVID-19 did not have a significantly higher burden of type 2 diabetes, while they had a significant trend to a lower burden of type 1 diabetes (48, 49) .

Our study has some limitations. First, our cohort consists of hospitalized patients with COVID-19, excluding asymptomatic or paucisymptomatic patients who were treated at home. Second, preadmission and postdischarge FBG were not available for all patients, and we cannot exclude a selection bias. Third, death might be an important competing risk (in the sense that prevent us from appreciating diabetes cases in patients who are deceased), and we did not perform a competing risk analysis. Fourth, we only systematically assessed FBG, and we acknowledge that more specific markers of beta cell function, such as serum insulin and or C-peptide levels, would have provided relevant information. Fifth, this has been a single-center study, covering mainly Europeans. Consequently, the conclusion should be limited to Europeans, and we cannot exclude that in other ethnic groups the COVID-19 encounter may led to a different clinical outcome in view of different beta cell functional capacity (ie, Asians as compared to Europeans). Finally, a 6-month follow-up may be too short to unveil a direct diabetogenic effects of the SARS-CoV-2 virus. In fact, there may be a delay between immunologic factors or infection damages and the onset of diabetes. It will be, therefore, crucial to conduct long-term follow-up studies in these patients.

COVID-associated dysglycemia is a complication of SARS-CoV-2 infection, and this clinical entity still needs to be adequately characterized in relationship with preexisting (pre) diabetes or glucose intolerance. It is clear from our study that patients with COVID-associated dysglycemia had increased levels of inflammatory markers and indicators of organ injuries and showed a poorer clinical outcome. This strongly support the need to screen all patients COVID-19 pneumonia for hyperglycemia at the time of admission, despite a mute personal or family history of diabetes, and treat their hyperglycemia promptly to achieve and maintain a good glycemic control during hospitalization. On the other hand, our data do not support the persistence of a long-term disruption of glycometabolic control after SARS-CoV-2 infection. Large epidemiological studies in the next years will be required to clarify whether COVID-19 induce permanent diabetes. However, at this time, any alarmist claim on an increased risk of diabetes after SARS-CoV-2 infection should be interpreted with caution.

Diabetes as a risk factor for greater COVID-19 severity and in-hospital death: a meta-analysis of observational studies

Burden of diabetes mellitus and its impact on COVID-19 patients: a meta-analysis of real-world evidence

Predictors of adverse prognosis in COVID-19: a systematic review and meta-analysis

Sanchis-Gomar F. Diabetes mellitus association with coronavirus disease 2019 (COVID-19) severity and mortality: a pooled analysis

Effects of hyperglycaemia on complications of COVID-19: a meta-analysis of observational studies

Acute kidney injury and kidney replacement therapy in COVID-19: a systematic review and meta-analysis

Registry of arterial and venous thromboembolic complications in patients with COVID-19

COVID-19 infection may cause ketosis and ketoacidosis

COVID-19 infection may cause ketosis and ketoacidosis

Diabetic ketoacidosis precipitated by Covid-19 in a patient with newly diagnosed diabetes mellitus

High prevalence of COVID-19-associated diabetic ketoacidosis in UK secondary care

Autoantibody-negative insulin-dependent diabetes mellitus after SARS-CoV-2 infection: a case report

A case of combined diabetic ketoacidosis and hyperosmolar hyperglycemic state in a patient with COVID-19

New-onset type 1 diabetes in children during COVID-19: multicenter regional findings in the U.K. Diabetes Care

Newly diagnosed diabetes and diabetic ketoacidosis precipitated by COVID-19 infection

De Sanctis V. Newly-onset type 1 diabetes mellitus precipitated by COVID-19 in an 8-month-old infant

Type 1 diabetes onset triggered by COVID-19

The association of diabetes with COVID-19 disease severity: evidence from adjusted effect estimates

SARS-CoV-2 cell entry depends on ACE2 and TMPRSS2 and is blocked by a clinically proven protease inhibitor

SARS-CoV-2 receptor angiotensin I-converting enzyme type 2 (ACE2) is expressed in human pancreatic β-cells and in the human pancreas microvasculature

A human pluripotent stem cell-based platform to study SARS-CoV-2 tropism and model virus infection in human cells and organoids

SARS-CoV-2 infects and replicates in cells of the human endocrine and exocrine pancreas

SARS-CoV-2 infection induces beta cell transdifferentiation

SARS-CoV-2 cell entry factors ACE2 and TMPRSS2 are expressed in the microvasculature and ducts of human pancreas but are not enriched in β cells

Expression of SARS-CoV-2 entry factors in the pancreas of normal organ donors and individuals with COVID-19

ACE2 expression in pancreas may cause pancreatic damage after SARS-CoV-2 infection

Association of blood glucose control and outcomes in patients with COVID-19 and preexisting type 2 diabetes

Association between acute pancreatitis and COVID-19: could pancreatitis be the missing piece of the puzzle about increased mortality rates?

Acute and long-term disruption of glycometabolic control after SARS-CoV-2 infection

Dysglycemia after COVID-19 pneumonia: a six-month cohort study

Biobanking for COVID-19 research

Early predictors of clinical outcomes of COVID-19 outbreak in

Dysglycemia after COVID-19 pneumonia: a six-month cohort study

COVID-19 survival associates with the immunoglobulin response to the SARS-CoV-2 spike receptor binding domain

Diabetes care in the hospital: standards of medical care in diabetes-2020

Data from: no evidence of long-term disruption of glycometabolic control after SARS-CoV-2 infection

Potential metabolic and inflammatory pathways between COVID-19 and new-onset diabetes

Antibody response to multiple antigens of SARS-CoV-2 in patients with diabetes: an observational cohort study

Proportion of newly diagnosed diabetes in COVID-19 patients: a systematic review and meta-analysis

Outcomes in patients with hyperglycemia affected by COVID-19: can we do more on glycemic control? Diabetes Care

New onset diabetes, type 1 diabetes and COVID-19

New-onset diabetes in COVID-19: time to frame its fearful symmetry

New-onset type 1 diabetes in children during COVID-19: multicenter regional findings in the U.K. Diabetes Care

The association between COVID-19 and type 1 diabetes mellitus: A systematic review

COVID-19 in people with diabetes: understanding the reasons for worse outcomes

Impaired glucose metabolism in patients with diabetes, prediabetes, and obesity is associated with severe COVID-19

COVID-19 and diabetes: the contributions of hyperglycemia

Risk of clinical sequelae after the acute phase of SARS-CoV-2 infection: retrospective cohort study

High-dimensional characterization of postacute sequelae of COVID-19

This work was funded by Program Project COVID-19 OSR-UniSR and Ministero della Salute (COVID-2020-12371617).Author Contributions: A.L.: data curation, investigation, writing (review and editing); A.C.: data curation, investigation, writing (review and editing); C.M.: data curation, investigation, writing (review and editing); A.M.: investigation; R.M.: investigation; R.N.: investigation; C.T.: investigation; P.R.Q.: data curation; F.C.: funding acquisition; V.L.: resources, formal analysis; E.B.: writing (review and editing); M.S.: resources, formal analysis, writing (review and editing); and L.P.: conceptualization, methodology, formal analysis, writing (original draft). L.P. is the guarantor of this work and, as such, had full access to all the data presented in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. The final manuscript has been read and approved by all named authors.

Correspondence: Lorenzo Piemonti, MD, Diabetes Research Institute, San Raffaele Scientific Institute, Via Olgettina 60, 20132 Milan, Italy. Email: piemonti.lorenzo@hsr.it.Disclosures: The authors have no conflict of interest to disclose in relation to the topic of this manuscript. Data Availability: Some or all data sets generated during and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.