New-Onset Diabetes and COVID-19: Evidence from a Global Clinical Registry

doi:10.21203/rs.3.rs-2405391/v1

Download PDF

Article

New-Onset Diabetes and COVID-19: Evidence from a Global Clinical Registry

https://doi.org/10.21203/rs.3.rs-2405391/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Mounting evidence shows association between COVID-19 and new diagnoses of diabetes. It is unclear, however, if COVID-19 increases detection of pre-existing diabetes or if it can induce new-onset of the disease.

Methods

We established a global online registry of COVID-19-related diabetes (CoviDIAB) using a web-enabled data capture system (Dendrite Clinical Systems). In this study we aimed to investigate whether COVID-19 can induce new-onset diabetes, its subtypes and clinical manifestations. To this end, we analyzed clinical and laboratory data from cases of newly-diagnosed diabetes occurring during or within four weeks from an episode of COVID-19. To exclude pre-existing hyperglycaemia, new-onset diabetes was defined as: blood sugar levels above diabetes thresholds (fasting glycaemia ≥ 126 mg/dL or non-fasting glycemia > 200 mg/dL), no prior history of the disease or use of glucose-lowering medications, and HbA1c < 6·5% at presentation.

Results

Between October 2020 and April 2022, 67 contributors from 61 hospitals in 25 countries entered data on 537 eligible cases of newly-diagnosed diabetes. New-onset diabetes was identified in 102 of 473 newly-diagnosed cases with recorded HbA1c (22%). Among adults, diabetes subtypes were type 2 (59%) and “not-yet known” (41%). There were two cases of new-onset type 1 diabetes among children. Hyperglycaemia persisted beyond resolution of the infection in 39 of 89 (45%) patients with new-onset diabetes who survived the episode of COVID-19. Further follow-up data beyond 3-months was available for 28 such cases, showing remission of diabetes in five and persistent diabetes in 23 cases (82%).

Conclusions

This study shows clinical plausibility for a diabetogenic effect of COVID-19, supporting screening for diabetes in people who contract the infection. Further investigation is warranted to confirm mechanisms of viral interference with glucose metabolism.

The CoviDIAB registry is accessible online at http://covidiab.e-dendrite.com.

Health sciences/Endocrinology/Endocrine system and metabolic diseases/Diabetes

Health sciences/Diseases/Infectious diseases

It is well established that people with diabetes have higher risk of severe complications and mortality from COVID-19.^1,2 On the other hand, new-onset hyperglycemia and acute metabolic complications of pre-existing diabetes have been observed during hospitalization for acute phase of COVID-19.^3–7 Data from various countries also show increased incidence of diabetes in the last two years of the pandemic.^8–10 Several studies have found increased risk of incident diabetes and use of anti-hyperglycemic drugs among people who had COVID-19 compared to people with no evidence of infection.^11,12

These observations suggest that COVID-19 may exacerbate pre-existing diabetes or even induce new onset of the disease. Given the high global prevalence of both conditions, understanding the link between COVID-19 and diabetes would have important implications for clinical practice, public health, and scientific research.

It is yet unclear, however, if the association between COVID-19 and diabetes reflects indirect health consequences of the ongoing pandemic (i.e., increased body weight as an effect of lockdowns, leading to insulin resistance),¹³ increased detection of pre-existing disease, or a direct diabetogenic effect of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the virus that causes COVID-19.

Increased detection of pre-existing disease could explain newly-diagnosed diabetes both during and after an episode of COVID-19. In fact. diabetes can go undetected for many years and about 46% of all global cases of diabetes in adults are estimated to be undiagnosed;¹⁴ hence, it is plausible that increased hospitalizations and medical tests during the pandemic of COVID-19 may alone reveal pre-existing disease.

Preliminary biological evidence, however, supports a possible viral influence on metabolic regulation. COVID-19 is associated with low-grade inflammation, which in turn, may induce or exacerbate insulin resistance.¹⁵ Several studies also suggest that SARS-CoV-2 can infect and replicate in insulin-producing pancreatic beta-cells and impair production and secretion of insulin.^16–18 Furthermore, SARS-CoV-2 enters human cells by binding to Angiotensin-converting enzyme 2 (ACE-2) receptors and other proteins that are expressed in various organs involved in glucose metabolism, including the adipose tissue, the liver, and the small intestine;^16–20 direct or indirect cell and tissue damage in such organs may plausibly lead to metabolic dysfunctions.

Whether such mechanisms or other unknown biological effects of SARS-CoV-2 can induce clinically relevant alterations of glucose metabolism is, however, still matter for debate.

Characterizing the clinical manifestations of COVID-19 related diabetes and the time-based relationship between the onset of hyperglycemia and infection could help unravel the question. The clinical manifestations of COVID-19-related diabetes, however, are still poorly characterized. Available data derive from case-reports, small single-center clinical series, or non-clinical databases of incident diabetes at the population level.^1–9 These studies do not provide sufficient information on key aspects of diabetes and patient’s history to assess whether observations of newly-diagnosed diabetes associated with COVID-19 reflect pre-existing disease or represent new-onset diabetes. Furthermore, the largest series reported so far relate to new diagnoses of diabetes occurring at distance of several months from the acute episode of COVID-19. Given the high incidence of diabetes in the population, these studies are not suitable to infer about cause-effect relationship.

Verification of true new-onset diabetes in closer temporal relationship with an acute episode of COVID-19 would provide a crucial piece of evidence regarding the plausibility of a diabetogenic effect of SARS-Cov-2.

We established a global registry of COVID-19 related diabetes (CoviDIAB Registry)²¹ and called on healthcare providers from around the world to share a minimal set of clinical and laboratory data about clinical observations of newly diagnosed diabetes and severe metabolic complications of pre-existing diabetes associated with COVID-19.

In this study of CoviDIAB data we sought to investigate whether COVID-19 can acutely induce new-onset diabetes and the clinical manifestations of this condition.

To do so, we studied cases of newly diagnosed diabetes occurring in close temporal association with COVID-19 (during or within four weeks from acute infection). To assess the potential of a direct diabetogenic effect of SARS-CoV-2, we specifically analyzed HbA1c data at time of diabetes presentation to rule out pre-existing hyperglycaemia and verify the new-onset of the disease.

Data Sources and Contributors

Between October 30, 2020 and April 30, 2022, 67 contributors from 61 hospitals in 25 countries (Supplementary Appendix, Tables S1 and S2) entered data on the 537 eligible cases of newly-diagnosed diabetes. Their geographical distribution was as follows: Middle East and North Africa (n = 170), South and Southeast Asia (n = 168), North America and Caribbean (n = 81), Europe (n = 71), Western Pacific Region (n = 31), Africa (n = 13), South and Central America (n = 3).

Cases Eligible for Analysis

At the time of this analysis a total of 897 cases were entered into the Registry. One hundred and two cases were deemed ineligible for analysis due to unclear diagnosis of either diabetes (n = 100) or COVID-19 (n = 2) (Fig. 1). Among 795 eligible cases, there were 258 cases (32%) of severe metabolic complications of diabetes and 537 cases of newly diagnosed diabetes (68%), which were the focus of this study.

Demographics and Clinical Characteristics of Newly Diagnosed Diabetes

Detailed data about 537 newly diagnosed cases of diabetes associated with COVID-19 are presented in Table 1. Of these, 418 were adults (age ≥ 20 years; 78%) and 119 were children/adolescents (age 0–19 years; 22%). The most represented ethnic groups among adults were South and Southeast Asian (42%), White-European (19%), and Arab (18%). Among adults, 277 patients were male (67%) and 141 female (33%). There were 61 males (51%) and 58 females (49%) among children/adolescents. Obesity, (defined by a BMI ≥ 30 kg/m² or ≥ 27·5 kg/m² for Asians) was present in 51% of adult patients.

Table 1

Patients reported with Newly Diagnosed Diabetes: Patient demographics, treatments, and outcomes
	All (n = 537)	Adults (n = 418)	Children/adolescents (n = 119)
Age, years	42 ± 22	51 ± 15	9 ± 4
BMI, kg/m²	27 ± 9	30 ± 8	18 ± 5
Female sex (%)	199 (37·1)	141 (33·4)	58 (48·7)
Ethnicity (%)
South East Asian	181 (33·7)	177 (42·3)	4 (3·4)
Arab	155 (28·9)	77 (18·4)	78 (65·5)
White	112 (20·9)	81 (19·4)	31 (26·1)
Other Asian	43 (8·0)	41 (9·8)	2 (1·7)
Black African	27 (5·0)	26 (6·2)	1 (0·8)
Hispanic or Latino	3 (0·6)	3 (0·7)	0 (0·0)
Other	3 (0·6)	3 (0·7)	1 (0·8)
Unknown	20 (3·7)	10 (2·4)	2 (1·7)
Covid-19 diagnosis (%)
RNA test	462 (86·0)	385 (92·1)	77 (64·7)
Antibody test	78 (14·5)	31 (7·4)	47 (39·5)
Radiological	366 (68·2)	345 (82·5)	21 (17·6)
Blood analytes
Fasting glycaemia, mg/dL	191 ± 83	187 ± 80	223 ± 100
Post-prandial glycaemia, mg/dL	312 ± 182	256 ± 145	456 ± 191
HbA1c, %	9·0 ± 2·9	8·3 ± 2·8	11·1 ± 2·4
Treatment (%)
Insulin only	307 (57·2)	201 (48·1)	106 (89·1)
Insulin + non-insulin medication	41 (7·6)	37 (8·9)	4 (3·4)
Metformin	33 (6·1)	33 (7·9)	0 (0·0)
SGLT2	3 (0·6)	3 (0·7)	0 (0·0)
DPP-4i	2 (0·4)	2 (0·5)	0 (0·0)
SU	2 (0·4)	2 (0·5)	0 (0·0)
GLP-1RA	1 (0·2)	1 (0·2)	0 (0·0)
Other	5 (0·9)	1 (0·2)	4 (3·4)
Non-insulin medication only	76 (14·2)	74 (17·7)	2 (1·7)
Metformin	62 (11·5)	60 (14·4)	2 (1·7)
DPP-4i	10 (1·9)	9 (2·2)	1 (0·8)
SU	9 (1·7)	9 (2·2)	0 (0·0)
GLP-1RA	6 (1·1)	6 (1·4)	0 (0·0)
TZD	5 (0·9)	5 (1·2)	0 (0·0)
SGLT2	1 (0·2)	1 (0·2)	0 (0·0)
No medication	113 (21·0)	106 (25·4)	7 (5·9)
Outcomes (%)
Admitted to ICU	117 (21·8)	96 (23·0)	21 (17·7)
Endotracheal intubation	74 (13·8)	67 (16·0)	7 (5·9)
Coma	13 (2·4)	2 (0·5)	11 (9·2)
Died	35 (6·5)	34 (8·1)	1 (0·8)
Data are mean ± SD or count (%)

Polyuria, polydipsia, and fatigue were reported in 15%, 13%, and 16% of adults vs 72%, 71%, and 5% of children/adolescents.

Diabetes Subtypes for Newly Diagnosed Diabetes

Diabetes subtypes, as diagnosed and reported by treating clinicians, were as follows: 60% type 2 diabetes (T2D); 24% type 1 diabetes (T1D), 15% yet unknown/undetermined and 1% latent autoimmune diabetes in adults (LADA). Specifically, T2D accounted for 76% of cases among adults (Fig. 2A) and 3% among children/adolescents (Fig. 2B). T1D represented 96% of pediatric cases and 4% of adult cases. The distribution of diabetes subtypes by geographic regions is shown in Fig. 3.

Verification of “New-onset Diabetes”

Cases of newly diagnosed diabetes were assessed for evidence of hyperglycemia preceding Covid-19 infection. To this end, we defined “new-onset diabetes” as: 1. Glycemia above diagnostic thresholds for diabetes (fasting glycemia ≥ 126 mg/dL or non-fasting glycemia > 200 mg/dL; 2. no prior history of diabetes or use of glucose-lowering medications; 3. HbA1c levels below diabetic range (< 6·5%) at presentation.

HbA1c data were available for 473 of 537 patients with newly diagnosed diabetes (Fig. 4). The overall distribution of HbA1c levels among these patients is shown in Fig. 5A.

There were 102 cases (22%) that met our definition of “new-onset diabetes” (HbA1c < 6·5%) and that also occurred during or within four weeks of an episode of COVID-19 (Fig. 5B). Of these, 98 were adults and 4 were children/adolescents. Mean HbA1c on admission/first observation for new-onset diabetes was 5·9% (2·7–6·4%; Table 2) whereas mean fasting and non-fasting glycemia were 160 ± 41 mg/dL (90–312 mg/dL) and 218 ± 88 mg/dL (96–750 mg/dL) respectively. Among adults, T2D accounted for 59% of cases whereas diabetes subtype was reported as “not yet known” in 41% (Fig. 5C). Among children/adolescents there were two cases with a diagnosis of T1D and two reported as “not yet known”.

Table 2

Patients with Confirmed New-Onset Diabetes: Demographics, treatments, and outcomes
	All (n = 102)	Adults (n = 98)	Children/adolescents (n = 4)
Age, years	50 ± 18	51 ± 16	13 ± 5
BMI, kg/m²	27 ± 5	27 ± 5	21 ± 5
Female sex (%)	36 (35·3)	33 (33·7)	3 (75·0)
Ethnicity (%)
South East Asian	71 (69·6)	70 (71·4)	1 (25·0)
Other Asian	20 (19·6)	19 (19·4)	1 (25·0)
White	6 (5·9)	6 (6·1)	0 (0·0)
Arab	4 (3·9)	2 (2·0)	2 (50·0)
Other	1 (1·0)	1 (1·0)	0 (0·0)
Covid-19 diagnosis (%)
RNA test	93 (91·2)	89 (90·8)	4 (100·0)
Antibody test	8 (7·8)	7 (7·1)	1 (25·0)
Radiological	88 (86·3)	86 (83·8)	2 (50·0)
Blood analytes
Fasting glycaemia, mg/dL	160 ± 41	159 ± 39	175 ± 103
Post-prandial glycaemia, mg/dL	218 ± 88	213 ± 71	333 ± 284
HbA1c, %	5·9 ± 0·5	5·9 ± 0·5	5·9 ± 0·1
Treatment (%)
Insulin only	28 (27·5)	25 (25·5)	3 (75·0)
Insulin + non-insulin medication	15 (14·7)	15 (15·3)	0 (0·0)
Metformin	14 (13·7)	14 (14·3)	0 (0·0)
DPP-4i	1 (1·0)	1 (1·0)	0 (0·0)
SU	1 (1·0)	1 (1·0)	0 (0·0)
Non-insulin medication only	40 (39·2)	39 (39·8)	1 (25·0)
Metformin	36 (35·3)	35 (35·7)	1 (25·0)
DPP-4i	6 (5·9)	6 (6·1)	0 (0·0)
GLP-1RA	6 (5·9)	6 (6·1)	0 (0·0)
SU	2 (2·0)	2 (2·0)	0 (0·0)
TZD	1 (1·0)	1 (1·0)	0 (0·0)
No medication	19 (18·6)	19 (19·4)	0 (0·0)
Outcomes (%)
Admitted to ICU	31 (30·4)	28 (28·6)	3 (75·0)
Endotracheal intubation	26 (25·5)	25 (25·5)	1 (25·0)
Coma	0 (0·0)	0 (0·0)	0 (0·0)
Died	16 (15·7)	16 (16·3)	0 (0·0)
Data are mean ± SD or count (%)

Sex, ethnic and BMI distribution of new-onset diabetes cases are presented in Table 2. Of note, 38% of T2D cases had no known predisposing factors for the disease (e.g., obesity, hypertension or dyslipidemia) (Fig. 5D). Moreover, there were three reported cases of acute renal injury and two cases of diabetic ketoacidosis, despite no previous history of either complication prior to COVID-19.

Outcomes of COVID-19 related New-Onset Diabetes

Admission to ICU, endotracheal intubation, and death among patients with new-onset diabetes were reported in 30%, 25%, and 16%, respectively (Table 2).

Among patients with new-onset diabetes who survived COVID-19 and for whom follow-up data were available (89 cases), hyperglycemia persisted beyond resolution of the infection in 39 cases (45%).

Further follow-up data beyond 3-months were only available for 28 such cases, showing remission of hyperglycaemia in five and persistent diabetes in 23 cases (82%) (Fig. 4).

This study analyzed clinical data from a global registry of 795 cases of diabetes associated with COVID-19, including 537 cases of newly diagnosed diabetes and 258 cases of acute metabolic complications of pre-existing diabetes. We specifically assessed cases of newly diagnosed diabetes for evidence of new onset of the disease in close temporal relationship with the episode of COVID-19. The results of this analysis provide clinical plausibility for a diabetogenic effect of COVID-19.

In this study, most patients who received a new diagnosis of diabetes in association with COVID-19 exhibited levels of HbA1c above thresholds used for diabetes diagnosis at presentation, suggesting that diabetes likely preceded the infection. This finding is not surprising, considering that about 46% of all global cases of diabetes in adults are undiagnosed.¹⁴

However, our study also shows that newly diagnosed diabetes associated to COVID-19 cannot be entirely explained by pre-existing disease. In fact, we were able to verify new-onset diabetes in at least 102 cases of newly diagnosed cases associated with COVID-19. These patients exhibited glycemic levels above thresholds for diagnosis of diabetes with HbA1c levels still in the non-diabetic range, consistent with onset of hyperglycemia during SARS-CoV-2 infection.

We have also documented persistence of hyperglycemia beyond the resolution of COVID-19 in at least 39 cases of new-onset diabetes with follow-up data available (45%).

We acknowledge that HbA1c levels above 6·5% at time of diagnosis may still be compatible with new-onset diabetes. Our strict definition of “new onset-diabetes” in this study may therefore underestimate the phenomenon; however, this approach reduces the likelihood of pre-existing disease. On the other hand, HbA1c between 6·0% and 6·5% could indicate that a state of pre-diabetes preceded the infection. However, acute progression from pre-diabetes to diabetes would still suggest a negative effect of COVID-19 on glucose metabolism.

In the aggregate, these results provide evidence of clinically relevant alterations of glucose metabolism acutely occurring with COVID-19 infection. Although these data do not prove that SARS-CoV-2 causes diabetes, the findings are the most meaningful clinical evidence to date in support of a possible diabetogenic effect of SARS-CoV-2. Confirmation of a cause-effect relationship between COVID-19 and diabetes will require further, consistent, and complementary evidence from epidemiologic, biologic, and clinical investigations.

While a potential viral etiology has been generally associated with T1D,²² our study shows that T2D is the dominant subtype among patients with new-onset diabetes in the acute phase of COVID-19. This finding is consistent with other reports showing that T2D also accounts for the majority of newly diagnosed cases of diabetes associated with the post-acute phase of COVID-19.^8,9,11,12

Also, a small clinical study using continuous glucose monitoring and hormone level assessment documented various alterations of glucose metabolism, including dysglycemia, insulin resistance and abnormal cytokine profile in association with COVID-19 infection, resembling pathophysiologic characteristics of T2D.²³

These observations suggest a complex pathophysiology, consistent with evidence that SARS-CoV-2 can infect not only the pancreas^16–18 but also other metabolic organs such as the adipose tissue, the liver and the small intestine.^19,20

The association between COVID-19 and T2D in this and other studies raises the intriguing hypothesis that other viruses, globally endemic over the last several decades, may have contributed to the recent epidemic of T2D. Although the hypothesis needs verification, it may provide new impetus and direction for the search of the elusive cause of T2D and the understanding of its pathophysiology.

Our analysis could only rule out pre-existing disease in two pediatric cases and one adult case of newly diagnosed T1D. However, we cannot exclude that this subtype may develop more often in association with COVID-19. In fact, definitive diagnosis of diabetes subtype was not feasible or available in 41% of adult cases with new-onset diabetes in this series. Moreover, this study was not designed to look at potential long-term consequences of COVID-19 on diabetes. Despite such caveats, our findings seem reassuring in regard to the previously hypothesized risk that COVID-19 may induce a large, epidemic wave of T1D.^8,24

Stress-induced, temporary hyperglycemia can occur with acute illnesses or corticosteroid therapy, especially pneumonia, both inherently associated with COVID-19.^14,25 It is therefore not possible to rule out that stress hyperglycemia might explain at least some of the observations of new-onset diabetes associated with an acute episode of COVID-19. In this study, however, hyperglycemia persisted beyond the acute episode of COVID-19 in about half (45%) of all cases of new-onset diabetes and durability of diabetes was documented for at least three months in more than 80% of cases where hyperglycemia survived the infection. These data suggest that COVID-19 can induce durable alterations of glucose metabolism. Our results are consistent with reports of persistent hyperglycemia in patients with newly diagnosed diabetes occurring during or after COVID-19^7,11,12 as well as with findings from other studies showing sustained alterations of glucose metabolism even in patients who do not develop diabetes during COVID-19.²³

Although about half (55%) of patients with new-onset diabetes in this study experienced remission of hyperglycaemia with the resolution of the infection, it remains unclear if these patients remain at risk of disease recurrence or development of new diabetes in the future, similarly to gestational diabetes.

Altogether, these observations warrant follow-up studies to understand the evolution of diabetes and the risk of future diabetes after COVID-19.

There are several strengths of this study. By collecting data of clinical observations of new-onset diabetes associated with COVID-19 from 25 countries, our study confirms this phenomenon across multiple world’s regions and ethnicities. The global outreach of CoviDIAB also allowed us to gather the largest series of clinically documented new-onset diabetes associated with COVID-19 or any viral infection to date. Furthermore, the close temporal association between SARS-CoV-2 infection and the onset of diabetes in this series makes the study more relevant for inference of cause-effect relationship.

This study has several limitations. As an international registry that relies on voluntary contribution from clinicians across many hospitals and countries, CoviDIAB data inherently reflect heterogeneity in practice and clinical judgement of contributing clinicians. Information about pathophysiology (e.g., C-peptide, antibodies, etc.) is limited, since these tests are not part of standard clinical practice in many countries. Hence, we could not independently confirm assessment of diabetes subtypes and must rely on clinical assessment by contributing clinicians.

Our follow-up data are also still limited. Most of the observations in this study relate to earlier phases of the pandemic and may not reflect effects of newer variants of SARS-CoV-2.

In conclusion, our study provides clinical plausibility for a diabetogenic effect of SARS-CoV-2. This finding warrants further research to confirm a possible etiologic role of SARS-CoV-2 in diabetes and to investigate the mechanisms of viral interference in glucose metabolism. Our data support screening for diabetes during the acute and post-acute phase of COVID-19.

Funding:

The CoviDIAB registry was established as a collaborative effort by King’s College London (London UK) and Monash University (Melbourne, Australia). King’s Health Partners and Monash University provided seed funds for the project. The study was also supported in part by the Technische Universität Dresden (TUD) - King's College London (KCL) Transcampus, Meydan Family Charitable Trust and Somnio Global USA and by an investigator-initiated grant from Novo Nordisk.

Acknowledgements:

We are indebted to Dr. Efstathia Papada and Dalal Qanaq at King’s College London (KCL) for their help reviewing online registration of contributors and verification of their institutional affiliations. We’d like to thank Dr Elizabeth Blackmore for her expert assistance with legal aspects of data sharing agreements. We thank Peter Walton and his team at Dendrite for their courtesy and their technical assistance with the set-up of the online platform of this registry.

We are also grateful to the following colleagues for their help with promotion of the registry and for encouraging or facilitating data-sharing: Dr. Hessa Al-Kandari, Prof. Mark Cooper, Dr. Laura McEwen, Prof. Stefano Del Prato, Prof. Barbara McGowan, Dr. David Hopkins, Prof. Mi Suk Lee, Dr. Mohammad Al Jumaah, Dr. Lior Tolkin, and Prof. Hamid Baradaran.

Disclosures:

Prof. Rubino reports receiving research funding from Medtronic, Johnson & Johnson and Novo Nordisk, consulting fees from GI Dynamics and lecturing fees from Johnson & Johnson, Medtronic, and Novo Nordisk. Prof. Rubino also serves on the data safety advisory board of GT Metabolic Solutions and is President of the Metabolic Health Institute (non-profit). Prof. Amiel has served on advisory boards for Novo Nordisk and Medtronic and has lectured at educational events sponsored by Novo Nordisk and Sanofi. She is a co-investigator on the EU IMI programme HypoRESOLVE and a member of the International Hypoglycaemia Study Group. Prof. Eckel reports receiving consulting fees from Amgen, Better Co., Lexicon, Novo Nordisk, The Healthy Aging Co., W.W. and Amarin, and fees for serving on the Arrowhead advisory board. Prof. Khunti has received consulting fees from AstraZeneca, Novartis, Novo Nordisk, Sanofi-Aventis, Eli Lilly, Merck Sharp & Dohme, Boehringer Ingelheim, Bayer, Berlin-Chemie AH / Menarini Group, Janssen and Napp, and is Chair of the Ethnicity Subgroup of the UK Scientific Advisory Group for Emergencies (SAGE). Prof. Mingrone reports receiving consulting fees from Novo Nordisk, Fractyl Inc. and Recor Inc., and has patents with Metadeq Inc. and GHP Scientific Ltd. Prof. Mbanya reports receiving fees for serving on advisory boards from Servier Laboratories and Novo Nordisk. Prof. Tan reports receiving grants from the Caswell Diabetes Institute, University of Michigan, and is a retiree of Eli Lilly Company and receives a monthly pension. Dr. Alessa reports receiving fees for serving on advisory boards and speaker honoraria from Sanofi, Novo Nordisk, Eli Lilly, Astra Zeneca, Servier, Boehringer Ingelheim and MSD, and has received support from Horizon for his attendance at ENDO 2022 and from Sanofi for his attendance at ATTD 2022. Dr. Alessa is a board member of the Gulf Association of Endocrinology and Diabetes (GAED).

Inclusion Criteria for Data Entry into the Registry

The CoviDIAB registry accepts data about “newly diagnosed diabetes” as well as “severe metabolic complications of pre-existing diabetes” associated with COVID-19.

Accepted methods of COVID-19 diagnosis, included antigen tests, antibodies tests or typical radiologic findings of COVID-19 related pneumonia.

For newly diagnosed diabetes, inclusion criteria included: 1. fasting glycemia ≥ 126 mg/dL or non-fasting glycemia > 200 mg/dL; 2. no prior history of diabetes and 3. no current or prior use of glucose-lowering drugs.

For severe metabolic complications of pre-existing diabetes inclusion criteria included acute onset of one or more of the followings: diabetic ketoacidosis, severe insulin resistance, hyperosmolar hyperglycemic state, and acute renal injury.

The present study focusses on the analysis of newly diagnosed diabetes. The study of metabolic complications of diabetes associated with COVID-19 addresses distinct research questions and will be the subject of future reports.

Modality of Data Collection

The registry draws data from across the world using a web-enabled data capture system (Dendrite Clinical Systems), which is accessible online at http://covidiab.e-dendrite.com.

The registry remains open-ended and is anonymized at the point of data entry. Data that could directly and indirectly reveal patients’ identities, (e.g., names, residential address, date of birth, hospital number, etc.) are not entered into the registry. To ensure maintaining anonymized data while allowing communication between investigators and data contributors as well as entry of follow-up data, a progressive unique number (Registry Unique Number, RUN) is generated for each patient. Only clinicians/authorized personnel at the contributing institutions can match the RUN with patient identifiers.

Data Contributors

Contributors register on the online platform and must provide information about their professional credentialing, institutional affiliation, type of medical practice and years of practice. Data about clinical observations are entered by clinicians who are members of the care-team or by authorized research personnel. Contributors must sign specific agreements and disclaimers and take full responsibility for entering reliable clinical information as well as respecting local laws and institutional regulations for data sharing and patients’ privacy.

Type of Data Collected

The database includes a minimum dataset of mandatory information for each case that includes basic demographics, glycemia and a positive diagnosis of COVID-19, plus a broader dataset. This includes other demographics, disease characteristics (including assessment of diabetes type by clinicians), co-morbidities, medication types and dosage, laboratory data, including measures of insulin, C-peptide, and islet antibody status where available, drugs used for the treatment of COVID-19 (i.e., dexamethasone, antivirals, etc.), diabetes outcomes and outcomes of COVID-19 (e.g., ICU usage, endotracheal intubation, death), setting of clinical observations and time-relationship between the diagnosis of COVID-19 and the first diagnosis of diabetes or its metabolic complications.

Data Management and Quality Control

The registry automatically screens and filters out inadequate or incorrect data submission. To this end, an initial set of questions blocks data entry if responses to key questions appear to conflict with one another. For instance, data entry for patients with newly diagnosed diabetes is halted by the system if contributors declare that there is a previous history of diabetes, or usage of medications for treating diabetes. Furthermore, data entry is manually checked by the research team looking for inconsistencies. Using RUN numbers, contributors were also individually contacted when necessary to verify and/or request data for variables of interest. The technology used for this registry is equipped with automated conversion of laboratory units. A descriptive analysis of data was performed at regular intervals to evaluate potential sources of bias. Cases were assessed for confirmed laboratory diagnosis of both COVID-19 and diabetes. Cases with reported levels of fasting and non-fasting glycemia below internationally accepted thresholds for diagnosis of diabetes were excluded from analysis.

Ethics

As an anonymized audit of standard clinical practice, this registry did not require Ethics approval in the UK, where the registry is located and data stored. However, individual contributors have obtained local ethics approval for sharing of data according to regulations and policies in their jurisdictions and Institutions.

Specific Aims and Methods of this Analysis

The present study focusses on the analysis of newly-diagnosed cases of diabetes associated with COVID-19. A descriptive analysis of demographics, diabetes type, glycemic levels, management, and outcomes was performed on all eligible cases.

To assess the potential of a direct diabetogenic effect of SARS-CoV-2, we specifically studied cases with close temporal association between COVID-19 infection and the onset of diabetes (within 4 weeks from COVID-19 diagnosis). These cases were assessed for evidence of hyperglycemia preceding Covid-19 infection. To this end, we defined “new-onset diabetes” as: 1. Glycemia above diagnostic thresholds for diabetes (fasting glycemia ≥ 126 mg/dL or non-fasting glycemia > 200 mg/dL; 2. no prior history of diabetes or use of glucose-lowering medications; 3. HbA1c levels below diabetic range (< 6·5%) at presentation.

Statistical analysis

Prior to analysis, all data were checked for possible data entry errors. Contributing clinicians were contacted to verify values where necessary. A descriptive analysis was conducted, data are presented as mean ± SD or count (%). Data were analyzed using SPSS version 28 (IBM Corporation, NY, USA) or GraphPad Prism Version 9·1 (GraphPad Software Inc. La Jolla, USA).

Singh AK, Gillies CL, Singh R, et al. Prevalence of co-morbidities and their association with mortality in patients with COVID-19: A systematic review and meta-analysis. Diabetes Obes Metab 2020; 22(10): 1915–24.
Barron E, Bakhai C, Kar P, et al. Associations of type 1 and type 2 diabetes with COVID-19-related mortality in England: a whole-population study. Lancet Diabetes Endocrinol 2020; 8(10): 813–22.
Chee YJ, Ng SJH, Yeoh E. Diabetic ketoacidosis precipitated by Covid-19 in a patient with newly diagnosed diabetes mellitus. Diabetes Res Clin Pract 2020; 164: 108166.
Sathish T, Kapoor N, Cao Y, Tapp RJ, Zimmet P. Proportion of newly diagnosed diabetes in COVID-19 patients: A systematic review and meta-analysis. Diabetes Obes Metab 2021; 23(3): 870–4.
Unsworth R, Wallace S, Oliver NS, et al. New-Onset Type 1 Diabetes in Children During COVID-19: Multicenter Regional Findings in the U.K. Diabetes Care 2020; 43(11): e170-e1.
Li H, Tian S, Chen T, et al. Newly diagnosed diabetes is associated with a higher risk of mortality than known diabetes in hospitalized patients with COVID-19. Diabetes Obes Metab 2020; 22(10): 1897–906.
Cromer SJ, Colling C, Schatoff D, et al. Newly diagnosed diabetes vs. pre-existing diabetes upon admission for COVID-19: Associated factors, short-term outcomes, and long-term glycemic phenotypes. J Diabetes Complications 2022; 36(4): 108145.
Barrett CE, Koyama AK, Alvarez P, et al. Risk for Newly Diagnosed Diabetes > 30 Days After SARS-CoV-2 Infection Among Persons Aged < 18 Years - United States, March 1, 2020-June 28, 2021. MMWR Morb Mortal Wkly Rep 2022; 71(2): 59–65.
Rathmann W, Kuss O, Kostev K. Incidence of newly diagnosed diabetes after Covid-19. Diabetologia 2022; 65(6): 949–54.
Kendall EK, Olaker VR, Kaelber DC, Xu R, Davis PB. Association of SARS-CoV-2 Infection With New-Onset Type 1 Diabetes Among Pediatric Patients From 2020 to 2021. JAMA Netw Open 2022; 5(9): e2233014.
Xie Y, Al-Aly Z. Risks and burdens of incident diabetes in long COVID: a cohort study. Lancet Diabetes Endocrinol 2022; 10(5): 311–21.
Rezel-Potts E, Douiri A, Sun X, Chowienczyk PJ, Shah AM, Gulliford MC. Cardiometabolic outcomes up to 12 months after COVID-19 infection. A matched cohort study in the UK. PLoS Med 2022; 19(7): e1004052.
Bakaloudi DR, Barazzoni R, Bischoff SC, Breda J, Wickramasinghe K, Chourdakis M. Impact of the first COVID-19 lockdown on body weight: A combined systematic review and a meta-analysis. Clin Nutr 2021.
Beagley J, Guariguata L, Weil C, Motala AA. Global estimates of undiagnosed diabetes in adults. Diabetes Res Clin Pract 2014; 103(2): 150–60.
Bornstein SR, Rubino F, Ludwig B, et al. Consequences of the COVID-19 pandemic for patients with metabolic diseases. Nat Metab 2021; 3(3): 289–92.
Muller JA, Gross R, Conzelmann C, et al. SARS-CoV-2 infects and replicates in cells of the human endocrine and exocrine pancreas. Nat Metab 2021; 3(2): 149–65.
Steenblock C, Richter S, Berger I, et al. Viral infiltration of pancreatic islets in patients with COVID-19. Nat Commun 2021; 12(1): 3534.
Wu CT, Lidsky PV, Xiao Y, et al. SARS-CoV-2 infects human pancreatic beta cells and elicits beta cell impairment. Cell Metab 2021; 33(8): 1565–76 e5.
Hikmet F, Mear L, Edvinsson A, Micke P, Uhlen M, Lindskog C. The protein expression profile of ACE2 in human tissues. Mol Syst Biol 2020; 16(7): e9610.
Berni Canani R, Comegna M, Paparo L, et al. Age-Related Differences in the Expression of Most Relevant Mediators of SARS-CoV-2 Infection in Human Respiratory and Gastrointestinal Tract. Front Pediatr 2021; 9: 697390.
Rubino F, Amiel SA, Zimmet P, et al. New-Onset Diabetes in Covid-19. N Engl J Med 2020; 383(8): 789–90.
Coppieters KT, Boettler T, von Herrath M. Virus infections in type 1 diabetes. Cold Spring Harb Perspect Med 2012; 2(1): a007682.
Montefusco L, Ben Nasr M, D'Addio F, et al. Acute and long-term disruption of glycometabolic control after SARS-CoV-2 infection. Nat Metab 2021; 3(6): 774–85.
Hollstein T, Schulte DM, Schulz J, et al. Autoantibody-negative insulin-dependent diabetes mellitus after SARS-CoV-2 infection: a case report. Nat Metab 2020; 2(10): 1021–4.
McCowen KC, Malhotra A, Bistrian BR. Stress-induced hyperglycemia. Crit Care Clin 2001; 17(1): 107–24.

Yes there is potential Competing Interest. Prof. Rubino reports receiving research funding from Medtronic, Johnson & Johnson and Novo Nordisk, consulting fees from GI Dynamics and lecturing fees from Johnson & Johnson, Medtronic, and Novo Nordisk. Prof. Rubino also serves on the data safety advisory board of GT Metabolic Solutions and is President of the Metabolic Health Institute (non-profit). Prof. Amiel has served on advisory boards for Novo Nordisk and Medtronic and has lectured at educational events sponsored by Novo Nordisk and Sanofi. She is a co-investigator on the EU IMI programme HypoRESOLVE and a member of the International Hypoglycaemia Study Group. Prof. Eckel reports receiving consulting fees from Amgen, Better Co., Lexicon, Novo Nordisk, The Healthy Aging Co., W.W. and Amarin, and fees for serving on the Arrowhead advisory board. Prof. Khunti has received consulting fees from AstraZeneca, Novartis, Novo Nordisk, Sanofi-Aventis, Eli Lilly, Merck Sharp & Dohme, Boehringer Ingelheim, Bayer, Berlin-Chemie AH / Menarini Group, Janssen and Napp, and is Chair of the Ethnicity Subgroup of the UK Scientific Advisory Group for Emergencies (SAGE). Prof. Mingrone reports receiving consulting fees from Novo Nordisk, Fractyl Inc. and Recor Inc., and has patents with Metadeq Inc. and GHP Scientific Ltd. Prof. Mbanya reports receiving fees for serving on advisory boards from Servier Laboratories and Novo Nordisk. Prof. Tan reports receiving grants from the Caswell Diabetes Institute, University of Michigan, and is a retiree of Eli Lilly Company and receives a monthly pension. Dr. Alessa reports receiving fees for serving on advisory boards and speaker honoraria from Sanofi, Novo Nordisk, Eli Lilly, Astra Zeneca, Servier, Boehringer Ingelheim and MSD, and has received support from Horizon for his attendance at ENDO 2022 and from Sanofi for his attendance at ATTD 2022. Dr. Alessa is a board member of the Gulf Association of Endocrinology and Diabetes (GAED).

SupplementaryMaterial.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

New-Onset Diabetes and COVID-19: Evidence from a Global Clinical Registry

Status:

Version 1

Abstract

Background

Methods

Results

Conclusions

Figures

Introduction

Results

Data Sources and Contributors

Cases Eligible for Analysis

Demographics and Clinical Characteristics of Newly Diagnosed Diabetes

Diabetes Subtypes for Newly Diagnosed Diabetes

Verification of “New-onset Diabetes”

Outcomes of COVID-19 related New-Onset Diabetes

Discussion

Declarations

METHODS

References

Additional Declarations

Supplementary Files

Status:

Version 1