Early in the pandemic, it became clear that Covid-19 infection spelt trouble for people with diabetes. While not at greater risk of catching the virus, they were more likely to be hospitalised and suffer complications. One study found that diabetes was the single biggest predictor of mortality (adjusted for age and sex) among hospitalised Covid-19 patients.

Along the way another, perhaps more surprising, pattern was emerging too. Doctors found that many Covid-19 patients, with no related prior history, were developing diabetic symptoms. In some cases, the symptoms resolved along with the infection, but in others, they persisted long after the patient had recovered. Researchers began to deliberate whether the link between Covid-19 and diabetes might run in both directions.

The strength of the evidence

One 2020 meta-analysis found that 14% of hospitalised Covid-19 patients developed new-onset diabetes. Another study suggested that rates of gestational diabetes had surged during the pandemic, while a retrospective analysis found that children who had been ill with Covid-19 were more likely to develop type 1 or type 2 diabetes.

In March 2022, the most compelling evidence to date was published in The Lancet Diabetes & Endocrinology. The study in question looked at over 180,000 veterans who had been diagnosed with Covid-19. Compared to a control group, Covid-19 patients had a 40% higher risk of developing diabetes over the subsequent 12 months. That equates to around 13 additional diagnoses for every 1,000 people.

Dr KM Venkat Narayan, a professor of medicine and endocrinology at Emory University School of Medicine , wrote an editorial in The Lancet to accompany this research. He thinks the link between Covid-19 and diabetes is strong enough to warrant further investigation.

“They study authors have done a good job and taken care of various confounding factors like age, BMI, gender, socioeconomic factors, among others,” he says. “Whether or not there is a direct causal link, this is a public health issue. Even small increases in the incidence of a chronic disease like diabetes would mean lots of new cases in the health system.”

How are the two connected?

So, can Covid-19 really trigger diabetes? And, per the suggestion of the Lancet study authors, should diabetes be considered a ‘component’ of long Covid? Narayan thinks that, while this is a “good working hypothesis”, we should rule out other factors before declaring the novel coronavirus as the culprit.

“As a scientist, I try to stay as agnostic as possible,” he says. “The people who develop a SARS CoV-2 infection and those who develop diabetes probably share various socioeconomic and geographical factors, so that could be a confounder. Does your lifestyle change when you have Covid-19, or are you under greater surveillance and consequently getting diagnosed with diabetes? These are the kinds of questions that need to be answered.”

Dr Kathleen Wyne, an endocrinologist at The Ohio State University Wexner Medical Center, remarks that any acute stress or severe illness can cause high blood glucose levels. On top of that, hospitalised patients with Covid-19 are treated with high-dose steroids, which raise sugar levels in anyone susceptible.

“Basically stress and/or steroids can bring out diabetes,” she says. “Perhaps diabetes is not ‘part of long-COVID’, and it is now occurring in people who were at risk for diabetes but may not have met the criteria so soon.”

That said, Narayan concedes there may be various causal mechanisms at play. According to one line of research, the novel coronavirus might infect pancreatic cells via a protein called ACE-2, affecting their ability to produce insulin. Other theories suggest the virus may trigger an excessive immune response, damaging pancreatic cells or that inflammation throughout the body could make tissues less responsive to insulin. 

While none of these theories have been proven– they are largely based on cell culture studies – researchers believe they are worth further exploration.

“What we really need are carefully designed epidemiological studies, combined with pancreatic biobanks where we can study the pancreases of people who unfortunately died of SARS CoV-2,” says Narayan.

How should these patients be treated?


Whatever lies behind the uptick in cases, it’s clear that health systems will need to find ways of dealing with the burden. Might these new-onset diabetes patients benefit from dedicated treatment strategies?

Narayan thinks that, while this is an important question, we can’t answer it satisfactorily until we better understand the causal mechanisms.

“If a Covid-19 infection has affected the beta cells of the pancreas, you might use the kind of drugs that are good at overcoming that problem,” he says. But he says we may not yet be at  the stage to come up with treatment guidelines specifically for this group. “It might be the same thing as regular diabetes, it might be different, we don’t know.”

Since we don’t yet have any specific studies on this group, Wyne suggests the best approach is to follow the guidelines for newly diagnosed diabetes patients.

“Last year the Glycemia Reduction Approaches in Diabetes (GRADE) study showed that the best agents for newly diagnosed type 2 diabetes are metformin and GLP-1 agonists,” she says. “While, sulfonylurea is not a preferred agent, based on cardiovascular and kidney risk, we may want to include an SGLT-2 inhibitor as soon as possible. If the diabetes is not well controlled then we would look to include insulin early.”

Screening for diabetes

One thing that can be done for Covid-19 patients specifically, is to test their glucose levels during the acute stage of infection and bring that under control if it is high. You could also test their glucose levels at follow-up appointments as a diabetes screening plan.  

“I think the most important thing here is to improve the guidelines for glucose testing,” says Narayan. He suggests doctors should perform a fasting glucose test or a hemoglobin A1C test, six months to a year after the original infection. A glucose tolerance test might also unearth any damage to the pancreas.

Wyne thinks we need to work harder to flag metabolic syndromes and pre-diabetes, and intervene as early as possible. That way, acute illnesses like Covid-19 will be less likely to trigger diabetes.

“If one has adequate pancreatic reserve then they can handle the stress and go back to normal glucose levels in three to six months,” she says. “But if the pancreas has already lost too much reserve – through genetics, age, stress, metabolic syndrome etc – then a Covid-19 infection will be the precipitating event , although the process was already well underway.”

For the time being, there is still a lot we don’t know about the connection between long Covid and diabetes. However, we do know that diabetes rates are rising – and this alone should sound warning bells for health systems.

 “As we come out of the pandemic, this is the time to be reflective, and say, how do we manage this dual burden?” says Narayan. “To me, the message here is that when a pandemic happens, you can’t ignore the other diseases.”