Abstract
Aims
To provide an overview of reported cases of new-onset type 1 diabetes mellitus (T1D) following COVID-19 infection.
Methods
PubMed and Scopus library databases were screened for relevant case reports published between January 2020 and June 2022. Study design, geographic region or language were not restricted.
Results
Twenty studies were identified and involved 37 patients (20 [54%] male, 17 [46%] female). Median age was 11.5 years (range 8 months–33 years) and 31 (84%) patients were aged ≤17 years. Most patients (33, 89%) presented with diabetic ketoacidosis (DKA). In total, 23 (62%) patients presented at the time of positive COVID-19 testing and 14 (38%) had symptoms consistent with COVID-19 infection or a previous positive test (1–56 days). Diabetes symptomatology was provided in 22 cases and (19, 86%) reported polyuria, polydipsia, polyphagia, fatigue, or weight loss or a combination of the aforementioned in the preceding weeks (3 days–12 weeks). Of the 28 patients that had data on acute and long-term treatment, all recovered well and most were managed with basal bolus insulin regimens. Quality assessment showed that most reports were either ‘good’ or ‘moderate quality’.
Conclusions
Although uncommon, new-onset T1D is a condition healthcare professionals may expect to see following a COVID-19 infection.
Introduction
In December 2019, an atypical case of viral pneumonia was discovered in Wuhan, Hubei Province of China. The cause was a novel coronavirus strain termed, Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2).1,2 Subsequently, coronavirus disease 2019 (COVID-19) has affected hundreds of millions of individuals, placed significant pressure on public health structures, and caused unprecedented socio-economic implications worldwide.3,4
Patients infected with SARS-CoV-2 exhibit a clinical presentation that commonly includes fever, dry cough and respiratory distress. 5 The presentation of the disease caused by the virus may also include non-respiratory manifestations such as vomiting, diarrhoea and abdominal pain, with the extent of organ involvement being unpredictable, especially long-term.6–9 Recently, published data suggest that SARS-CoV-2 could also potentially contribute to an increase in the incidence of cases of type 1 diabetes mellitus (T1D), one of the most prevalent endocrine diseases in childhood. 10 Indeed in the UK, investigators reported an 80% increase in the number of cases of T1D in children in 2020 compared with previous years, which may suggest an association between COVID-19 infection and T1D presentation. 11 Previous research has found that seasonal viruses can be trigger factors for T1D.12,13 Therefore, the seasonal variation in the incidence of T1D may be partly explained by the periodic variation of viral infections. 14 Similarly, COVID-19 may precipitate/accelerate the presentation of T1D or, possibly contribute to the diagnosis of previously unrecognized T1D cases as patients present as diabetic emergencies. 15 The aim of our systematic review was to screen all currently available published literature and provide a complete summary of reported cases of T1D following COVID-19 infection.
Methods
Two independent [DS, KKT] searches were performed of PubMed and Scopus databases for studies published from 01 January 2020 to 30 June 2022 that provided data for new onset or exacerbation of diabetes following COVID-19 infection. Search terms included: “(COVID-19 OR COVID19 OR SARS-COV2) AND (Diabetes OR Diabetes mellitus OR DM)”. In addition, the reference lists of all included studies were checked for any potential additional publications. The study design, geographic region or language were not restricted; studies were included if they provided individual patient data. However, review articles, abstracts submitted in conferences and non-peer reviewed sources were excluded. Studies on in vitro and animal models were also excluded. The two reviewers [DS, KKT] first screened titles followed by abstracts and selected relevant studies. They independently extracted data with any disagreement settled by a third reviewer [KSK].
The following items, when available, were extracted: first author; publication date; region; gender; age; comorbidities; family history of diabetes; presenting symptoms; history of COVID-19 infection; laboratory test results; imaging findings; therapeutic management (acute and long-term).
Independent assessment of risk of bias (RoB) was performed by two authors [KSK, KKT] using the critical appraisal checklist by the Joanna Briggs Institute (JBI). 16 For case studies, the assessment is based on the reporting of eight different elements: patient demographics; medical history; health status; diagnosis; intervention; post-intervention health status; unanticipated events; takeaway lessons. Depending on the availability of information for every element, the studies were scored as follows: yes; no; unclear; not/applicable.6,16
Descriptive statistics were used to describe the demographics and clinical characteristics of the included patients. Means for continuous variables and frequencies and percentages for binary variables were used. The systematic review was conducted according to PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) regulations. 17 Ethical committee approval was not required and the protocol was register in the PROSPERO online database (CRD42022338271).
Results
The initial literature search yielded 564 publications. Following removal of duplicates and the exclusions, 20 studies were eligible for the systematic review (Figure 1).18–37 Seven of the studies were conducted in Europe, five in USA, four in North Africa, and four in Asia (Table 1). In terms of design, two studies were case series and 18 were case reports

Prisma flowchart demonstrating the screening process. IPD: individual patient data.
Characteristics of included studies.
Abbreviations: M, male; F, female; SC, subcutaneous; T1D, type 1 diabetes; T2DM, type 2 diabetes mellitus; NR, not recorded; DKA, Diabetic Ketoacidosis; LRTI, lower respiratory tract infection
In total, there were 37 cases of new onset T1D. Of these, 23 (62%) patients presented at the time of positive polymerase chain reaction (PCR) test, or rapid antigen COVID-19 testing. The remaining 14 (38%) patients had symptoms consistent with COVID-19 infection or had a positive COVID-19 test previously (i.e., 1–56 days prior to diagnosis of T1D) (Table 1).
Of the 37 patients, 20 (54%) were male and 17 (46%) were female. Median age was 11.5 years (range 8 months to 33 years) and 31 (84%) patients were aged ≤17 years. Details of medical history were provided for 34 patients and most (29, 85%) had no co-comorbidities. Co-morbidities included obesity (2, 5%), coeliac disease (1, 3%), gestational diabetes (1, 3%) and hypothyroidism (1, 3%). Information about relevant family history was provided for 24 patients, of whom 12 (50%,) had at least one family member with a history of either type 1 or type 2 diabetes (Table 1).
Data regarding onset of diabetes symptomatology was provided in 22 cases with the majority (19, 86%) reporting polyuria, polydipsia, polyphagia, fatigue, or weight loss or more typically a combination of the above in the preceding weeks (i.e., 3 days to 12 weeks). In the three (14%) remaining cases, the patients reported that they had no common clinical features of diabetes prior to diagnosis.
Most patients (33/37, 89%) presented with diabetic ketoacidosis (DKA). The most common symptoms were fever (19/37, 51%), weight loss (18/37, 49%), nausea/vomiting (15/37, 41%), osmotic symptoms (16/37, 43%), fatigue (13/37, 35%) and abdominal pain (12/37, 32%) (Table 1). With the exception of two cases where data were not provided,24,27 all patients had levels of haemoglobin A1c (HbA1c) >6.5%. For the 28 cases where blood glucose was recorded, all patients had blood levels >250 mg/dl (Table 2).
Laboratory results and imaging findings.
Abbreviations: Glu, glucose; BG, blood glucose; BE, base excess; HCO3; bicarbonate; HbA1c, haemoglobin A1c; NR, not recorded; IAA, insulin autoantibodies; ICA, islet-cell antibodies; ZnT8, Zinc transporter 8; HLA, Human Leukocyte Antigen; CT, computed tomography; CTX, chest X ray;
Diabetes autoantibodies (i.e., islet cell cytoplasmic autoantibodies (ICA), insulin autoantibodies (IAA), glutamic acid decarboxylase antibodies (GADA), insulinoma-associated-2 autoantibodies (IA-2A) and Zinc Transporter 8 autoantibodies (ZnT8A) were reported for 18 patients, of whom 13 (72%) had at least one positive result (Table 2). Results of human leukocyte antigen (HLA) typing were only provided in three cases. Similarly, imaging results were only available from a few patients (9, 24%) who underwent computed tomography (CT) scans or chest X rays (Table 2).
Data on treatment in the acute and long-term settings were provided for 28 patients (Table 1). In the majority of cases, IV fluids and insulin were the mainstay of treatment in the acute setting (26/28, 93%); the remaining two (7%) patients received subcutaneous insulin.18,22 For long-term treatment, all 28 patients received long term insulin regimens and one patient also received metformin.
Quality assessment of the eligible studies using the JBI critical appraisal tool for case studies, showed that most reports could be graded as either ‘good’ or ‘moderate quality’ (Table 3). However, only one report attained a perfect score. 29 Most reports only achieved 50% of the quality criteria and three reports20,27,31did not meet at least 50% of the criteria and were classified as high risk of bias. Across the 20 reports, the domain most commonly assessed as ‘not available’ was Q7 (unanticipated events during the admission).
Quality assessment of the included studies.
Q1: Were patient’s demographic characteristics clearly described? Q2: Was the patient’s history clearly described and presented as a timeline? Q3: Was the current clinical condition of the patient on presentation clearly described? Q4: Were diagnostic tests or methods and the results clearly described? Q5: Was the intervention(s) or treatment procedure(s) clearly described? Q6: Was the post-intervention clinical condition clearly described; Q7: Were adverse events (harms) or unanticipated events identified and described? Q8: Does the case report provide takeaway lessons?
● = Yes; ○ = No; ᴓ = Unclear; ◌ = Not Available.
Discussion
COVID-19 infection is a disease associated with a plethora of clinical manifestations. Emerging evidence suggests a potential link between COVID-19 and the development or exacerbation of T1D.10,11,15 In this systematic review, we comprehensively examined literature published from 01 January 2020 to 30 June 2022 to provide a current overview of reported cases of T1D following infection with SARS-CoV-2. We identified 20 reports, that included data from 37 patients for whom T1D had been reported at the time or following a COVID-19 infection. There were slightly more male than female patients and most were aged 17 years or younger. A majority presented with DKA, which was managed acutely with IV fluids and insulin therapy. For patients where long-term data were provided, they recovered well and most were managed with basal bolus insulin regimens. Our findings are consistent with previous systematic reviews on COVID-19 infection and T1D and, other autoimmune disorders such as Graves’ disease and nephrotic syndrome, that also observed a male dominance.38–40 In addition, most of our patients were young, and, as previously reported, DKA is more common in children. 41
Type 1 diabetes results from the autoimmune destruction of beta cells in the pancreas with genetic and environmental factors precipitating the disease onset. 42 Viruses, such as influenza, parainfluenza, human herpesvirus 6 and enteroviruses are considered as triggers of autoimmune T1D in genetically susceptible individuals and are known to precipitate diabetic emergencies, such as DKA in patients with pre-existing diabetes, or even unmask latent disease.43,44 DKA is more common in children and is caused by total or partial insulin deficiency, which favours the production of ketones and requires urgent treatment.41,45 The preclinical period that precedes disease onset is long and it has been suggested that infectious pathogens may enhance self-antigen presentation, resulting in the involvement of different autoantigens.46,47
Evidence of the impact of SARS-CoV-2 on new-onset T1D is conflicting. For example, one study conducted in the UK in 30 children up to the age of 16 years with new-onset T1D and DKA, found an evident increase during the pandemic and their results were based on confirmation of infection with testing or evidence of exposure. 48 Similarly, a 12-month cross sectional study conducted in the USA from March 2020 to March 2021, found a significant increase in DKA at the time of T1D diagnosis and a 57% increase from the previous year in cases among children. 49 In addition, a metanalysis on the risk of T1D and DKA in paediatric patients with T1D between the COVID-19 pre-pandemic and pandemic periods, concluded that during the first year of the COVID-19 pandemic the incidence of worldwide paediatric new-onset T1D, DKA, and severe DKA increased by 10%, 25%, and 20%, respectively. 50 In a study from Tunisia, that investigated the incidence of new onset DKA between March 2018 and February 2022, investigators reported a 48% increase during the pandemic with the frequency of T1D and T2D rising by 50% and 40% respectively. 51 By contrast, a study from Belgium found that the prevalence of anti-SARS-CoV-2 antibodies in people of the same age with new-onset T1D was similar to that found in children without diabetes. 52 Nevertheless, retrospective data from 27,292,879 patients from a national US database (Cerner Real-World Data) showed that COVID-19 diagnosis was associated with a significantly increased risk of new-onset T1D with specific ethnicities being more susceptible. 53 The investigators also reported that in patients with pre-existing T1D, the risk of developing DKA was significantly increased following COVID-19 diagnosis.
Reduced access to care during the pandemic may be a contributing factor related to the increase in cases with new-onset T1D presenting as DKA. In addition, lifestyle changes during the lockdowns, which included reduced physical activity and weight gain, could have also contributed to the increase in diabetes and DKA. However, SARS-CoV-2 infection may affect the pathogenesis of autoimmune diabetes. For example, it is well recognised that angiotensin converting enzyme 2 (ACE-2) is the entry point of SARS-CoV-2 to several organs including the pancreas, which supports the theory of direct injury. 54 ACE-2 is downregulated after endocytosis of SARS-Cov2, which subsequently leads to increased action of angiotensin II, a putative mechanism that could contribute to abnormal insulin secretion.55,56 Moreover, SARS-CoV-2 antibodies may trigger inflammatory destruction of beta-cells. 57 Cytokines that play an integral role in the course of the infection, increase the ACE-2 expression in beta-cells and thus their susceptibility to the virus. 57 Interestingly, interleukin-6 is increased in both DKA and COVID-19 and has been suggested as a potential prognostic factor. 58 There is some evidence to suggest that following initial insulin dependency, some patients that present with new onset diabetes and COVID-19 are adequately controlled on oral antihyperglycemic medications several months later, suggesting a clinical picture more in line with ketosis prone diabetes.59,60 However, data from our systematic review of 20 case reports involving 37 patients showed that HbA1c levels were increased in all patients and most tested positive for COVID-19 shortly before, or at the time of admission, and so we can assume that SARS-CoV-2 was a precipitating factor of disease onset similar to other viral infections rather than a causative factor.
To the best of our knowledge, our study is the first to construct a systematic review of the association between COVID-19 and onset or exacerbation of T1D. Our findings from 20 case reports, categorised as low risk of bias and satisfactory assessment based on case report quality criteria, are consistent with previous literature. However, there were some limitations. For example, only two databases were screened. In addition, papers were only included if they provided individual patient data. In addition, details on demographic and clinical characteristics were insufficiently reported in approximately half of the 20 studies and three did not meet >50% of the quality criteria and were considered high risk of bias. Therefore, more studies are required. Indeed, to investigate and identify a potential link between COVID-19 infection and the development or exacerbation of T1D, prospectively designed and large cohort studies are required together with epidemiological studies using extensive data from the years following the pandemic.
In summary, COVID-19 may accelerate the presentation of T1D or possibly unmask previously undiagnosed cases in a similar manner to other seasonal viruses.53,54 New onset T1D is possibly a condition physicians and other healthcare professionals may expect to see in patients following COVID-19 infection and the early recognition of the disorder is important for the optimal management of these patients.
Supplemental Material
sj-pdf-1-imr-10.1177_03000605231210403 - Supplemental material for COVID-19 induced type 1 diabetes: A systematic review of case reports and series
Supplemental material, sj-pdf-1-imr-10.1177_03000605231210403 for COVID-19 induced type 1 diabetes: A systematic review of case reports and series by Dimitra Stathi, Konstantinos Katsikas Triantafyllidis, Marina Zafeiri, Janaka Karalliedde and Konstantinos S. Kechagias in Journal of International Medical Research
Footnotes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
