SGLT-2 inhibitors are now recommended as first-line therapy for people with type 2 diabetes who have had a previous cardiovascular event. However, there are concerns over a few safety issues associated with these relatively new medications. A new analysis of recent clinical trials, published in a recent issue of Diabetologia, concludes that, overall, the benefits of SGLT-2 inhibitors outweigh the risks.

Around half of those with type 2 diabetes die from cardiovascular disease (CVD) but, in the last decade, there has been new hope in the form of SGLT-2 inhibitors. These drugs have been investigated in a number of cardiovascular outcome trials (CVOTs), including the CANVAS Program, CREDENCE, DECLARE-TIMI and EMPA-REG OUTCOME, in which they have been associated with reduced cardiovascular events and all-cause mortality.

These findings have led to SGLT-2 inhibitors being recommended in the latest international consensus as second-line therapy following the failure of metformin and as first-line therapy in those who have had a previous cardiovascular event. Set against this clinical benefit, however, there are the adverse effects that have been linked to the SGLT-2 inhibitors. These include genital and urinary-tract infections, volume depletion, diabetic ketoacidosis (DKA), bone fractures and cancer.

The cardiovascular efficacy of the SGLT-2 inhibitors in the CVOTs was estimated using data that are correctly powered and at a low risk of bias. Their safety, however, was estimated from observational studies, which have a higher risk of bias. There have been some meta-analyses, which have provided safety assessments, but safety and efficacy data have never been combined to show if there is actually a net clinical benefit from using SGLT-2 inhibitors in people with diabetes. Accordingly, Guillaume Grenet and colleagues, in Lyon, France, have carried out a quantitative estimation at low risk of bias of the risk/benefit ratio of the SGLT-2 inhibitors. The findings will help clinicians and people with type 2 diabetes make informed choices over whether to use these medications in the management of their condition.

Gathering evidence on SGLT-2 inhibitors

The literature was searched for randomised controlled CVOTs published between the end of 2014 and 14th September 2021. Studies that recruited only individuals with type 2 diabetes, with or without other conditions, were included. Trials comparing either placebo or active controls were included, but those comparing different SGLT-2 inhibitors were excluded. The primary outcomes looked for in the analysis were overall mortality and the risk-benefit ratio between the key efficacy outcomes and the primary safety outcomes. The efficacy outcomes were major adverse cardiovascular events (MACE) – that is, death from cardiovascular causes, non-fatal myocardial infarction and non-fatal stroke – and hospitalisation for heart failure (HHF). Cases of end-stage renal disease (ESRD) were also recorded. The primary safety outcomes were amputation, DKA and genital infection. These outcomes were chosen for their relevance to clinical practice. A number of secondary outcomes, like volume depletion, urinary infections, kidney problems, fracture and cancer were also recorded. 

For each outcome, the researchers extracted the sample size, the number of events in each arm of the trial, the follow-up time and the measurement of the treatment’s effect. A total of five double-blind placebo-controlled international CVOT trials, covering 46,969 participants, was included in the primary analysis. Two of the trials tested canagliflozin (CANVAS and CREDENCE), one tested dapagliflozin (DECLARE-TIMI), one tested empagliflozin (EMPA-REG OUTCOME) and one tested ertugliflozin (VERTIS-CV). Mean follow-up ranged from 2.6 years to 4.2 years.

When it came to the participants, mean age ranged from 63.0 to 64.4 years, mean duration of type 2 diabetes was 10.5 to 15.8 years and mean HbA1c was 65 to 67 mmol/mol. Mean body mass index (BMI) was 30.6 to 32.1 kg/m2 and mean estimated glomerular filtration rate (eGFR) was 56.2 to 85.3 ml/min per 1.73m2. Participants with a history of CVD ranged from 40.6 to 99.2%, while those with a history of heart failure ranged from 10.1 to 24% and those with a history of amputation from 1.4 to 5.3%.

Risk-benefit ratio of SGLT-2 inhibitors

There were 3303 deaths during the trials, with an overall mortality rate of 21 per 1000 person-years. Taking SGLT-2 inhibitors is associated with a 14% decreased risk of overall mortality compared with controls. Put another way, for 1000 individuals treated over 3.5 years, SGLT-2 inhibitors will decrease the number of deaths, on average, from 70 to 61.

A total of 4798 participants had a MACE, and SGLT-2 inhibitors reduced the risk by 9%, which means that if 1000 participants are treated for 3.5 years, the number of MACE will be reduced from 37 to 26.

Finally, 310 participants developed ESRD, and SGLT-2 inhibitors reduced this risk by 33%. So, if 1000 participants are treated for 3.5 years, the number of cases of ESRD will be reduced from 8 to 6.

The mean spontaneous rate of amputation was four per 1000 person-years. This means that SGLT-2 inhibitors are not associated with a significant increase in the risk of amputation, compared with controls. However, they were associated with a significant increase in DKA and in genital infections. This means that if 1000 participants were treated over 3.5 years, the number of cases of DKA would increase from 1 to 3 and the number of genital infections from 15 to 51. 

In summary, the findings of the CVOT trials suggest that if 1000 individuals are treated for 3.5 years, on average 22 of them will benefit, with nine MACE, 11 HHF and two cases of ESRD prevented. However, 38 will be harmed by SGLT-2 inhibitors – namely, two cases of DKA and 36 genital infections. This means, according to the analysis, that 778 people will not experience any of the following outcomes: MACE, HHF, ESRD, amputation, DKA or genital infection. The authors conclude that the balance between MACE and DKA and MACE and amputation remains in favour of SGLT-2 inhibitors being of benefit (by reducing MACE more than they increase DKA or amputation). For HHF, the benefit again outweighs the risk of DKA or amputation. However, the number of genital infections exceeded the number of MACE and HHF.

When it came to secondary outcomes, there were no significant differences in the risk of bladder, breast or kidney cancer, or fracture with SGLT-2 inhibitors compared with controls. There was, however, an increase in the risk of volume depletion.

In summary…

In a large population of individuals with type 2 diabetes and a high risk of CVD, the cardiovascular and renal benefits of SGLT-2 inhibitors were substantial, despite the risk of DKA and the hypothetical risk of amputation. This study puts actual numbers on the benefits in terms of MACE, HHF and ESRD versus the risks in terms of DKA and genital infections (see above). It should also be noted that this new analysis finds that SGLT-2 inhibitors reduce mortality by around 14% compared with controls, which is comparable to the 16% reduction over five years found in a 2021 study by Palmer et al.

To expand a little on the safety issues, the study did confirm an increased risk of DKA, genital infections, urinary tract infections and volume depletion for people on SGLT-2 inhibitors. This is consistent with the findings of a recent meta-analysis. Genital infections generally occur in the first months of treatment and are usually benign or mild and easy to treat. Of more concern is the potential increased risk of amputation, which was seen in the CANVAS trial. This might be related to canagliflozin, although the CREDENCE trial did not find this increased risk. Or it might even be a false positive. In conclusion, this new analysis, which covers several CVOT trials, does not find an increased risk of amputation, which is reassuring.

The authors go on to say that it’s important that the clinical relevance of the benefits of SGLT-2 inhibitors should be set against those of the risks. Most genital infections can be easily treated, whereas with stroke, myocardial infarction and heart failure there is a high risk of sequelae or even death. Their analysis also did not capture the risk of other microvascular complications, like retinopathy, which might occur with SGLT-2 treatment – this could, of course, be investigated in future studies.

Overall, though, this is a valuable analysis, showing that for individuals with type 2 diabetes and a high CVD risk, the cardiovascular and renal benefits seem to outweigh the risks. It is to be hoped that these results will help inform both clinicians and people with type 2 diabetes when discussing what to expect when starting on an SGLT-2 inhibitor.

To read this paper, go to: Marilly E, Cottin J, Cabrera N, Cornu C, Boussageon R, Moulin P, Lega JC, Gueyffier F, Cucherat M, Grenet G. SGLT-2 inhibitors in type 2 diabetes: a systematic review and meta-analysis of cardiovascular outcome trials balancing their risks and benefits. Diabetologia 4 August 2022.

For more on this topic, enrol on the following EASD e-Learning courses:

Any opinions expressed in this article are the responsibility of the EASD e-Learning Programme Director, Dr Eleanor D Kennedy.

Modern lifestyles tend to preclude a prolonged fast during the night, with food being consumed over a long time period, which may have a detrimental effect on metabolic health. A new study, reported in a recent issue of Diabetologia, shows that eating within a time window of ten hours can help improve glycaemic control in type 2 diabetes.

Round-the-clock availability of food, irregular sleep-activity patterns and exposure to artificial light characterise modern society. These factors can lead to a disrupted day-night rhythm, which may increase the risk of type 2 diabetes, as has been seen in shift workers. Most people in Western society spread their daily food intake over a minimum period of 14 hours, meaning they do not experience a prolonged fast during the night. Time-restricted eating (TRE), where food intake is restricted to a time window of 12 hours or less during the day, restores the nocturnal fast.

Research has shown that TRE leads to a number of metabolic benefits in people living with overweight or obesity. These include increased lipid oxidation, decreased plasma glucose and improved insulin sensitivity. However, the studies performed so far involved relatively short eating time windows, such as six to eight hours, which are not readily applicable in everyday life. And there has only been one study looking at TRE in type 2 diabetes, which used a nine-hour window.

There may be some weight loss with TRE, which naturally leads to improvements in metabolic health. But improvements may also occur in the absence of weight loss, indicating that some additional mechanisms come into play. Disturbed circadian rhythmicity in various metabolic processes may play a role in type 2 diabetes and may be brought on by a disturbed eating-fasting cycle – that is, eating too late or over too long a time period. Accordingly, TRE may improve metabolic health by extending the fasting period. Specifically, this could involve hepatic glycogen, which acts as a fuel at night when glucose levels should be low. A decrease in hepatic glycogen might lead to an increase in insulin sensitivity, which would, of course, benefit people with type 2 diabetes.

Patrick Schrauwen and colleagues, of Maastricht University Medical Center, The Netherlands, have tested this hypothesis with a three-week study of a ten-hour TRE programme, in free-living conditions, in a group of adults with type 2 diabetes.

The TRE experiment

This was a randomised crossover study consisting of two three-week intervention periods separated by a washout period of four or more weeks and involving 14 adults with type 2 diabetes. In the TRE intervention, participants consumed their habitual diet within a ten-hour window during the day, with their final meal completed no later than 18.00 hours. In the control intervention, they spread their habitual diet over at least 14 hours. Interstitial glucose was measured every 15 minutes with a Freestyle Libre Pro and fasted glycogen was measured on one occasion during each intervention. Various other biochemical analyses were carried out, including measurement of insulin sensitivity.


Hepatic glycogen content did not differ significantly between the TRE and control groups. Nor did hepatic and peripheral insulin sensitivity. However, there were some significant changes between the two groups. There was a larger insulin-stimulated non-oxidative glucose disposal and lower carbohydrate oxidation in the TRE group, which reflects a greater ability to form glycogen – namely, glucose uptake is directed more towards storage than oxidation.

When it came to the continuous glucose monitoring (CGM) measurements, mean 24-hour glucose levels were lower in the TRE group than in the control group (6.8 mmol/l versus 7.6 mmol/l). Also, the TRE group spent more time in the normal glucose range – 15.1 hours versus 12.2 hours – and less time in the high glucose range at 5.5 hours versus 7.5 hours. Furthermore, there were no differences between the two groups for time spent in hyperglycaemia, in the low glucose range, or in hypoglycaemia.

The researchers also looked at fasting plasma metabolites on day 20 and day 21 of each intervention. On day 20, the TRE group had had an overnight fast of 14 hours, and the control group 10 hours. Plasma glucose was lower in the TRE group at 7.6 mmol/l versus 8.6 mmol/l. Insulin, triglycerides and non-esterified fatty acids (NEFAs) were similar between the two groups. And on day 21, when fasting time was the same in both groups, fasting glucose was still lower in the TRE group at 8.0 mmol/l versus 8.9 mmol/l, and there were still no differences between the two in the other plasma metabolite levels.

TRE for type 2 diabetes?

So, this study looked at restricting food intake to a ten-hour time frame during the day for three weeks in a group of people with type 2 diabetes. The findings showed an increase in non-oxidative glucose disposal in the TRE intervention, which suggests the hoped-for increased glycogen storage was occurring. This, in turn, may help regulate 24-hour and postprandial glucose levels, which again would be a good result in type 2 diabetes management. Indeed, the study did find that 24-hour glucose levels were improved following TRE, driven mainly by decreased nocturnal glucose.

This study did not find an improvement in insulin sensitivity, in contrast to an earlier study which did. This might have been because this latter study involved a shorter, six-hour TRE regimen with earlier consumption of the final meal at 15.00, as opposed to 18.00 in the current study. So it would be worthwhile exploring the impact of different lengths of eating windows on metabolic health. Of course, the shorter the eating window, the harder it is to put into practice in everyday life, so there is a trade-off to be taken into consideration here.  

Besides an improvement in 24-hour glucose, TRE also improved time spent in the normal glucose range and decreased time spent in high glucose – both of which are clinically relevant in type 2 diabetes. And there was no increase in hypoglycaemia with TRE and no other adverse events, suggesting that adopting a ten-hour eating window is both safe and effective in type 2 diabetes.

The mechanisms underlying the improvements seen with TRE require further investigation, although the improvements in glucose control could be at least partly explained by weight loss.

The authors note that some participants were on glucose-lowering medication, which might have limited the impact of TRE. It would have been interesting to look at TRE in people not on medication, but then the findings would not have been representative of the general type 2 diabetes population. Another limitation is that the study duration was only three weeks. Given that significant benefits were seen even in this short time scale, a longer-term study is now needed.

In conclusion, a ten-hour TRE regimen for only three weeks decreases glucose levels and prolongs time spent in normoglycemia in comparison with spreading food intake over 14 hours. These findings suggest that TRE could play a realistic role in improving type 2 diabetes management. 

To read this paper, go to: Andriessen C, Fealy C, Veelen A, van Beek SMM, Roumans KHM, Connell NJ, Mevenkamp J, Moonen-Kornips E, Havekes B, Schrauwen-Hinderling VB, Hoeks J, Schrauwen P. Three weeks of time-restricted eating improves glucose homeostasis in adults but does not improve insulin sensitivity: a randomised crossover trial. Diabetologia 25 July 2022.

For more on type 2 diabetes management, enrol on the following EASD e-Learning course:

Any opinions expressed in this article are the responsibility of the EASD e-Learning Programme Director, Dr Eleanor D Kennedy.

A new review, reported in a recent issue of Diabetologia, looks at glycaemic thresholds for the appearance of the signs and symptoms of hypoglycaemia in people with and without type 1 diabetes. These appear at lower glucose levels for people with diabetes and, also of note, the thresholds for appearance of autonomic and neuroglycopenic symptoms are similar, contrary to popular belief. Dr Susan Aldridge reports.

Hypoglycaemia is an ongoing challenge for many people with type 1 diabetes as a consequence of their treatment with insulin. Therefore, any new research that sheds light on the underlying physiological and biochemical processes is welcome, as it may lead to ways of avoiding hypos or making them easier to manage.

When glucose levels fall into the hypoglycaemic range, a counterregulatory hormonal response is triggered. First, there is suspension of insulin production by the beta cell, followed by the release of several hormones – namely glucagon, adrenaline, noradrenaline, cortisol and growth hormone. Then we get autonomic (warning) symptoms, which hopefully will lead to a behavioural response – usually, ingesting fast-acting carbohydrate. The autonomic symptoms include sweating, anxiety, tremor, palpitations and feeling hot and tingling. These are usually thought to precede neuroglycopenic symptoms like difficulty speaking, confusion, dizziness, irritability, blurred vision and drowsiness, which are a sign that the brain is being starved of glucose.

The impact of varying degrees of hypoglycaemia on counterregulatory response can be studied by a method called the hyperinsulinaemic clamp. Insulin is infused into a participant to cause glucose levels to fall, alongside a variable glucose infusion titrated against frequent glucose measurements to achieve a state of hypoglycaemia at various predefined plateau values. At each plateau, hormone concentrations and symptom scores can be recorded to define the glucose levels at which responses occur.

We know there is variability both between and within individuals of the thresholds where these responses occur. They are also influenced by factors such as prior exposure to hypoglycaemia and hyperglycaemia, age and duration of diabetes. Factors specific to type 1 diabetes can influence counterregulatory and symptom responses so the corresponding glycaemic thresholds differ from those for people without diabetes. To explore this, Clementine Verhulst from Radboud University Medical Centre, Nijmegen, The Netherlands and colleagues in Denmark and the UK, carried out a systematic literature review of hyperinsulinaemic clamp studies in people with, and without, type 1 diabetes.

They looked at 63 papers, published between 1980 and 2018, involving 1332 participants. Eleven included only people with type 1 diabetes, 26 only those without diabetes and 26 included both.

Glycaemic thresholds for hormone responses and symptoms

The glycaemic thresholds for eliciting hormone responses all occurred at lower median glucose levels in people with type 1 diabetes than in those without diabetes – 3.4 versus 3.8 mmol/l for adrenaline, 3.0 versus 3.2 for noradrenaline, 2.8 versus 3.5 for cortisol and 3.2 versus 3.8 for growth hormone. For glucagon, the threshold for people without diabetes was 3.8 mmol/mol (this was not reached for people with diabetes). Neither duration of diabetes nor glycaemic control measured by HbA1c was associated with the glycaemic threshold level for any of the hormones measured.

The glycaemic threshold for the appearance of both autonomic and neuroglycopenic symptoms did not differ, occurring at a median of 3.0 mmol/l in people with diabetes and 3.4 mmol/l in those without diabetes. And, as found for the hormone responses, neither duration of diabetes nor HbA1c was associated with the threshold level of symptom responses in people with diabetes.

Implications of this study

This review shows that for people without diabetes, hormone responses to hypoglycaemia occur between 3.2 and 3.8 mmol/l, while lower glucose levels were required to elicit symptom responses. Hormone and symptom responses occur at lower glucose levels in people with type 1 diabetes.

Based on hypoglycaemic glucose clamp studies dating back to the 1980s, it has been assumed that the physiological response to hypoglycaemia in people without diabetes occurs at glucose levels below 3.9 mmol/l, with the release of glucagon and adrenaline, while symptoms occur at levels below 3.3 to 3.5 mmol/l. This new study suggests that both hormone and symptom responses actually occur at much lower glucose levels on average, in people with and without diabetes.

In people with type 1 diabetes, glycaemic thresholds occur at lower levels (0.2 to 0.7 mmol/l and 0.4 mmol/l lower for hormone and symptom responses, respectively) than they do for people without diabetes. There is a wide range of thresholds, though, for people with diabetes, suggesting greater variability within this population. This could be for a number of reasons – different prior exposure to hypos, loss of glucagon secretion, blunted adrenaline and noradrenaline responses and impaired hypo awareness. In fact, this study did find that those with impaired awareness had lower thresholds for appearance of autonomic symptoms and release of adrenaline. This is in line with other research which finds that recurrent hypos, which often occur in those with impaired awareness, shifts glycaemic thresholds for hormone responses to lower values.

The study also found that strict glycaemic control – low HbA1c – does not have an impact on glycaemic thresholds. This is important, for low HbA1c, resulting from strict glycaemic control, is often assumed to be associated with greater exposure to hypos and higher HbA1c with less exposure. This new finding suggests that advising people to relax their HbA1c target in order to avoid hypos won’t actually work, as the glycaemic threshold remains the same.

Another finding challenges conventional wisdom around hypos – that the threshold for autonomic and neuroglycopenic symptoms is actually the same. It’s generally been assumed, though, that autonomic symptoms, like sweating and palpitations, appear before neuroglycopenic symptoms, such as difficulty in thinking or speaking – that is, symptoms appear in a hierarchy. This assumption may have arisen because autonomic symptoms are more immediately obvious. Neuroglycopenic symptoms might only become apparent if someone is engaged in a task requiring information processing. This is important when it comes to educating people about hypos, because they are usually told about this hierarchy, when this study shows that it does not actually exist.  

Finally, what does this new study tell us about the definition of hypoglycaemia? The American Diabetes Association (ADA) defines hypoglycaemia non-numerically as ‘all episodes of an abnormally low plasma glucose concentration that expose the individual to potential harm’. What actually constitutes a hypo is still under debate – and is important to people with type 1 diabetes. The findings of this new study acknowledge that glucose levels below which hormone and symptom responses appear show high intra- and interindividual variability – and this should be acknowledged in this ongoing debate.

Finally, the authors caution that their analysis was based on data obtained during experimental hypoglycaemia. Evidence from spontaneous hypoglycaemia occurring in everyday life is also needed to complement discussions on how hypos are to be defined. 

In conclusion, these new findings may inform clinical practice when supporting and educating people with type 1 diabetes about hypos. Hopefully, they will also inspire further research into the detailed understanding of hypoglycaemia and inform discussion on its definition. 

To read this paper, go to:

Verhulst C, Fabricius T, Teerenstra S, Kristensen PL, Tack CJ, McCrimmon RJ, Heller S, Evans ML, Amiel S, Pedersen-Bjergaard U, de Galan BE, Hypo-RESOLVE consortium. Glycaemic thresholds for counterregulatory hormone and symptom responses to hypoglycaemia in people with and without type 1 diabetes: a systematic review. Diabetologia. 22 July 2022.

For more on this topic, enrol on the following EASD e-Learning course:

See also Professors Stephanie Amiel and Rory McCrimmon’s contribution to our series ‘The long and the short of it’, ‘Cognitive function and hypoglycaemia’.

Any opinions expressed in this article are the responsibility of the EASD e-Learning Programme Director, Dr Eleanor D Kennedy.

Previous research has shown that both type 1 and type 2 diabetes are associated with increased morbidity and mortality from COVID-19 infection. However, we don’t know if this increased risk arises from diabetes itself or its associated comorbidities, such as obesity and heart disease – or whether both are important. A new analysis, reported in a recent issue of Diabetologia, from the French CORONADO study of people hospitalised for COVID-19, shows that diabetes is an independent risk factor for a worse prognosis. Dr Susan Aldridge reports.

Early on in the COVID-19 pandemic, diabetes was found to be associated with an increased risk of severe outcomes, including death. Several studies subsequently confirmed this, including a whole-population analysis from England, showing an increase of COVID-19 related mortality, and a greater risk of COVID-19- related death or intensive care unit admission in Scotland – both studies comparing people with and without diabetes. But it’s not known whether this increased risk relates to diabetes itself, or to its comorbidities – or maybe both.

While there have been studies looking at the impact of single comorbidities, none has looked at how the burden of multimorbidity might affect COVID-19 outcomes in diabetes. However, there has been a review of observational studies that suggests that those whom the authors define as having a ‘more severe course’ of diabetes, and therefore more comorbidities, have a poorer COVID-19 prognosis compared with those who have ‘a milder course’ of diabetes.

To address the need to clarify the relationship between diabetes, multimorbidity and COVID-19 outcomes, the CORONADO team in France has carried out a dedicated study.

The CORONADO study

The Coronavirus SARS-CoV-2 and Diabetes Outcomes (CORONADO) study is a French nationwide study, with both retrospective and prospective data collection, looking at the phenotypic characteristics and prognosis of people with diabetes admitted to hospital with COVID-19 between 10 March and 10 April 2020. As such, it was well set up for carrying out this new study, whose objective was to determine whether diabetes is a prognostic factor for COVID-19, independent of age and diabetes-associated comorbidities. The diabetes cohort was matched 1:1 for age, sex and date of admission to hospital with participants who did not have diabetes. The researchers then compiled a measure known as the Charleson comorbidity index (CCi) for each participant, from their medical records. The CCi is used to capture comorbidities associated with mortality risk and, for the purposes of the study analysis, two versions were used – with diabetes either included or excluded as a comorbidity. The primary composite outcome set for the study was invasive mechanical ventilation (IMV) and/or death within 7 days or 28 days.

Diabetes and COVID-19 outcomes

There were 2210 people with diabetes matched with 2210 people without the condition in this study. Most had type 2 diabetes and, compared with the matched individuals, they were more likely to have obesity, hypertension, dyslipidaemia and/or cardiovascular disease, as you’d expect, and so their CCi, with and without diabetes included, was higher. And, on admission, the individuals with diabetes had higher blood glucose levels and lower estimated glomerular filtration rates (eGFRs).

Invasive mechanical ventilation and/or death within seven days occurred in 29.0% of the diabetes group compared with 21.6% of those without diabetes, and in 34.8% and 28.4%, respectively, within 28 days. Each individual outcome – IMV and death – occurred more frequently in those with diabetes.

A statistical analysis of the data showed that diabetes is independently associated with worse COVID-19 prognosis irrespective of comorbidity burden in a population hospitalised for the infection in France during the first wave of the pandemic. These findings add to other research showing that diabetes confers susceptibility to infectious disease, particularly influenza and pneumonia. Such susceptibility was also reported during the 2009 H1N1 influenza pandemic and in the more recent MERS-CoV outbreak.

Explaining the impact of diabetes on morbidity and mortality from COVID-19 is beyond the scope of this paper, but is urgently in need of further investigation. It’s been suggested that hyperglycaemia might play a role. The nationwide survey report from Scotland did find an association between HbA1c and COVID-19 severity, but the authors didn’t find that in the current study. And the study of the UK population found an increased risk of COVID-19 death in people with diabetes irrespective of their HbA1c level. However, the current study did notice a link between higher plasma glucose at admission and COVID-19 severity. The higher glucose could merely be a result of infection, or it may interfere with the immune response, worsening the outcome.

The authors note that their study was carried out during the first wave of the pandemic. Prevention and treatment of COVID-19 have improved since then. Further studies are needed to see if this has also improved outcomes, specifically for those with diabetes. Also, the population studied was severely ill, so we don’t know if the worse outcomes for those with diabetes would also be observed in the general population.

Unfortunately, as we know, COVID-19 has not gone away. As healthcare systems gear up to deal with further waves of infection during the autumn and winter months, these findings should be used to plan and target specific interventions to reduce COVID-19 morbidity and mortality among those with diabetes.

To read this paper, go to: Cariou B, Wargny M, Boureau A-S, Smati S, Tramunt B, Desailloud R, Lebeault M, Amadou C, Ancelle D, Belkau B, Bordier L, Borot S, Bourgeon M, Bourron O, Cosson E, Eisinger M, Gonfroy-Leymarie C, Julia J-B, Marchand L, Meyer L, Seret-Begeue D, Simon D, Sultan A, Thivolet C, Vambergue A, Vatier C, Winiszewski P, Saulner P-J, Bauduceau B, Gourdy P, Hadjadj S and on behalf of the CORONADO investigators. Impact of diabetes on COVID-19 prognosis beyond comorbidity burden: the CORONADO initiative. Diabetologia online 15 June 2022.

For more on the CORONADO initiative, watch Professor Samy Jadjadj’s contribution to our series Diabetes and COVID-19, Answering questions of risk.

Any opinions expressed in this article are the responsibility of the EASD e-Learning Programme Director, Dr Eleanor D Kennedy.

Revealing the transcriptome of each pancreatic islet cell type could take us a long way towards clarifying the underlying mechanisms of type 2 diabetes. Single-cell RNA sequencing offers this level of genetic information and the opportunities and challenges posed by this emerging technology are discussed in a recent paper in Diabetologia. Dr Susan Aldridge reports.

The study of islet cells is central to advancing our understanding of the pathophysiology of type 2 diabetes. The best known of the islet cells is the beta cell, which produces insulin. But there are also alpha, delta, PP and ghrelin cells, producing glucagon, somatostatin, pancreatic polypeptide and ghrelin respectively. What role do they play in type 2 diabetes?

Work on animal models has certainly improved our understanding of islet biology, as have experiments on human islets from organ donors. And now we have a new technology for exploring islet pathophysiology – single-cell RNA sequencing (scRNAseq). As the name suggests, scRNAseq allows the profiling of individual cell transcriptomes. It allows for transcriptome profiling of rare cell populations and detailed study of gene expression. Most tissues and cell populations, including islet material, are heterogeneous and scRNAseq allows the characterisation of this heterogeneity. 

Other advantages of scRNAseq include defining how distinct cell populations respond to stimuli and predicting cellular progression in, for instance, the path from progenitors to mature lineages. 

However, scRNAseq is still in its infancy. Most protocols currently capture only 5 to 20% of the transcripts in a cell. And in the current state of computational analysis, its output needs to be complemented by biological knowledge for validation. Nevertheless, scRNAseq has the potential to have a significant impact on our understanding of disease processes in complex tissues like pancreatic islets. Nils Wierup and Mtakai Ngara, of Lund University Diabetes Centre, have reviewed key scRNAseq studies on human and mouse islets that have been carried out in the last five years and discuss how these have increased our understanding of islet biology.  

Islet donors without diabetes

So far, there have been 19 scRNAseq studies on human islets, seven of which compared samples from donors with, and without, type 2 diabetes. These have shown that the technology is technically feasible in human islets and that the different protocols have various pros and cons – a good foundation for future research. The studies have generated road maps of cell type-specific gene expression with potentially valuable information on receptors, transcription factors and cell surface markers. The authors note that perhaps the most significant contribution is the characterisation of the rarer islet cell types – namely delta, PP and ghrelin cells – which has not previously been possible. 

When it comes to beta cells, it’s long been known that they are not all the same, and some of these scRNAseq studies have revealed subpopulations of these cells. However, we don’t yet know if these arise from donor differences – so this apparent heterogeneity should be carefully validated.  

Other highlights include a study that found 1188 beta cell genes associated with obesity and another showing differences in alpha and beta cell gene expression according to age, implying incomplete differentiation in juvenile donors. There was also evidence that alpha cells proliferate at five times the rate of beta cells from the same donor. 

We don’t know how pancreatic cells are renewed, but one of the scRNAseq studies revealed subpopulations of pancreatic duct cells as potential multipotent cells. This is supported by work that showed that when these cells were transplanted into mice, differentiation into all pancreatic lineages was observed – an important insight into islet cell biology. 

Islet donors with type 2 diabetes

One study noted differential expression (DE) – upregulation and downregulation – of 76 genes in beta cells, 100 genes in alpha cells and five genes in PP cells from type 2 donors. Further analysis showed genes responsible for mitochondrial energy metabolism and protein synthesis were downregulated, whereas apoptosis and cytokine-signalling genes were upregulated. Another study found as many as 1368 type 2 diabetes-associated genes in beta cells, many of which were affected similarly by type 2 diabetes and obesity. 

The ultimate goal of these studies, comparing cells from donors with and without type 2 diabetes, was to identify disease mechanisms in beta cells. In three studies, systematic global DE analysis was carried out. But when the authors compared these findings with their own dataset, they found no differentially expressed gene in common across the four studies that could be focused on for further research. And they further note that only one gene, FXYD2, has been convincingly replicated between studies. One interpretation might be that altered FXYD2 expression is a major cause of beta cell dysfunction in type 2 diabetes. We know that FXYD2 knockout mice show improved glucose tolerance, beta cell hyperplasia and elevated fasting and postprandial plasma insulin. But the authors point out that FXYD2 expression doesn’t actually seem to be genetically associated with either type 2 diabetes risk or beta cell function, as would be expected if altered expression of this gene was of major importance. 

So, an alternative interpretation is that current scRNAseq approaches do not really capture type 2 diabetes disease biology. If they did, you would expect genes with proven roles in beta cell pathophysiology to appear in DE work and to be replicated between studies. 

The authors suggest several reasons for this apparently disappointing observation. The inherently low detection rate of scRNAseq, mentioned above, could mean that disease biology is being missed. Also, studies are still small in terms of cell counts and individuals, leading to low statistical power. It is also becoming increasingly apparent that type 2 diabetes is a heterogeneous disease. Beta cells from an insulin-resistant individual will likely have different gene expression from one with a primary beta cell defect. Finally, suboptimal methods are being used in DE analysis and the data lacks adjustment for donor variance. 

In summary, studies currently find no association between type 2 diabetes and either cell-type composition or subpopulations of islet cells. Individual reports have put forward novel type 2 diabetes-associated genes, some of which have been functionally validated. But few have been replicated between studies.

The way forward

The scRNAseq studies described above, and others, have certainly increased our knowledge of islet cell biology. But, the authors say, the contribution they have made to our understanding of type 2 diabetes has so far been limited. This is because DE analysis has proved to be rather a blunt tool for identifying actual differences between two groups in these small-scale studies with heterogeneous starting material. A recent study has highlighted these shortcomings – for instance, identifying a bias towards highly expressed genes, but maybe missing others of significance? The authors are calling for better computational tools to use the full potential of scRNAseq technology. The ultimate test of such tools would be their ability to reproduce known patterns of gene expression. Fortunately, there are a number of approaches that fulfil this requirement and actually exploit cellular heterogeneity rather than averaging it, as is done in DE analysis. 

Another limitation is that most approaches used to analyse scRNAseq data overlook the possibility that gene expression might actually be organised and synchronised in genetic programmes – maybe for orchestrating specific cellular functions. An approach known as network analysis can deal with this – for example, by assessing co-expression of mRNAs within a population of cells. Network analysis has been shown to be applicable to identifying disease-relevant pathways and regulatory networks. In particular, differential network analysis (diNA), which enables detection of changes in the interplay between mRNAs rather than assessing changes in single mRNAs, has been shown to outperform DE analysis. Although diNA has not yet been applied to scRNAseq data, the authors suggest that combining this approach with a reliable method for addressing donor variation could be one way forward for exploiting the potential of scRNAseq technology. 

In addition, increased numbers of both samples and sequenced cells are needed to increase the analytical power of scRNAseq,  to increase the diversity of donors to allow for studies of the different subtypes of diabetes. One way would be to integrate existing datasets into the analysis. But, going forward, large-scale studies with uniform sampling and analysis protocols with freshly isolated islets from donors well characterised with clinical data, medical records and genome sequencing, will clearly be needed. 

In conclusion, scRNAseq has provided new insights into islet cell biology through revealing cell-type-specific gene expression in all islet cell populations. But when it comes to increasing our understanding of type 2 diabetes pathophysiology, the technology has not reached its full potential. However, the ongoing rapid development of scRNAseq technology with respect to analysis tools and protocols, along with decreased costs, promises much in terms of our understanding of the altered biology of each islet cell type in type 2 diabetes.

To read this paper, go to: Ngara M, Wierup N. Lessons from single-cell RNA sequencing of human islets. Diabetologia online 28 April 2022. 

For more on this topic, enrol on the following EASD e-Learning modules:

Any opinions expressed in this article are the responsibility of the EASD e-Learning Programme Director, Dr Eleanor D Kennedy.

Obstructive sleep apnoea, with its various health consequences, is more common among people with type 2 diabetes than those without the condition. A new study, reported in a recent issue of Diabetologia, finds that those with type 1 are also more at risk of this common sleep disorder, with autonomic neuropathy possibly playing a role, independent of obesity. Clinicians should therefore be on the alert for obstructive sleep apnoea as yet another potential complication of type 1 diabetes. Dr Susan Aldridge reports.

Obstructive sleep apnoea (OSA) is a common disorder characterised by repeated complete or partial upper airway obstruction during sleep. This leads to repeated episodes of oxygen saturation and desaturation, changes in heart rate, blood pressure and sympathetic activity, along with disruption of sleep architecture. Consequences include an increased risk of road traffic accidents, reduced workplace productivity, hypertension, coronary heart disease, impaired quality of life and increased mortality.

And OSA is also a well-established risk factor for type 2 diabetes. Recently, it’s been shown that those who already have type 2 diabetes are also at increased risk of developing OSA, suggesting a bi-directional relationship. Furthermore, cohort studies have shown that OSA in people with type 2 diabetes is associated with micro- and macrovascular complications.

However, little is known about OSA in type 1 diabetes. It’s certainly plausible that people living with type 1 are at risk of OSA because of the increasing prevalence in this population of obesity and the high prevalence of autonomic neuropathy, both of which contribute to OSA. Therefore, Ziyad Alshehri at the University of Birmingham, UK, and colleagues, have carried out a cohort study to assess whether people with type 1 diabetes are at an increased risk of incident OSA when compared with matched controls without diabetes. They also looked at factors influencing the risk of developing OSA in the type 1 diabetes cohort.

The researchers drew upon The Health Improvement Network (THIN), a primary care database, using records from January 1995 to January 2018. From this, they identified 34,147 people with type 1 diabetes who were matched to 129,500 people of the same sex, age and body mass index (BMI). The study population was a ‘typical’ type 1 diabetes population – young, poor glycaemic control, low prevalence of obesity and with relatively few prescriptions of cardiovascular medications.

OSA – another diabetes complication

By the end of the study, OSA had been diagnosed in 219 (0.64%) of the type 1 diabetes cohort and in 531 (0.41%) of the controls, over a median follow-up of 5.43 years. The median diabetes duration in those with type 1 diabetes who were diagnosed with 0SA was 19.36 years, which is approximately double the duration for those with type 1 who did not develop OSA.

Analysis adjusting for factors that could influence the risk of OSA like age, BMI and smoking and drinking status, revealed that people with type 1 diabetes had a 51% higher risk of developing OSA than those without the condition. Further analysis showed that male sex, age, being overweight or obese, use of lipid-lowering or blood pressure-lowering medication and a history of atrial fibrillation or depression all increased the risk of incident OSA.

This is the first study to look at the longitudinal relationship between type 1 diabetes and OSA. It finds that the risk of developing OSA is similar to that found in type 2 diabetes. At first, this might seem a bit surprising, given the younger age and lower prevalence of obesity in type 1. But there could be another mechanism at play here, other than age or obesity – both of which are known risk factors for OSA. The authors suggest that diabetic autonomic neuropathy might be a contributing factor. This is supported by the fact that OSA was more common among those with a longer duration of type 1 and that atrial fibrillation is a risk factor – for diabetic autonomic neuropathy is associated with both.


As mentioned above, OSA in type 2 diabetes is associated with a higher risk of micro- and macrovascular complications. Further research should look at whether this is also the case with OSA and type 1 diabetes. In addition, people with type 1 diabetes run an increased risk of hypoglycaemia and, if they also develop OSA, it means they may be at added risk of road traffic accidents. There are also, of course, implications for driving licences.

This new study also found that depression was a risk factor for OSA with type 1 diabetes. This echoes findings from studies in people without diabetes, which finds that depression is common in people with OSA. The good news is that treatment of the OSA with continuous positive airway pressure improves depressive symptoms.

Atrial fibrillation (AF) was another risk factor found in the current study and other research has found that up to 85% of those with AF do also have OSA, confirming the association found here. So, for all these reasons, clinicians looking after people with type 1 diabetes should suspect OSA among those with any of the above risk factors. It is a potential complication that should not be ignored and treatment should be put in place as soon as possible. 

To read this paper, go to: Alshehri Z, Subramanian A, Adderley N, Gokhale KM, Karamat MA, Ray CJ, Kumar P, Nirantharakumar K, Tahrani AA. Risk of incident obstructive sleep apnoea in patients with type 1 diabetes: a population-based retrospective cohort study. Diabetologia 24 May 2022.

Any opinions expressed in this article are the responsibility of the EASD e-Learning Programme Director, Dr Eleanor D Kennedy.

A new study in a recent issue of Diabetologia compares intermittent (flash) glucose monitoring with finger-prick testing in a group of pregnant women with type 1 diabetes. While flash did improve glycaemic control for a while, this was not sustained, and it was also linked to an increased risk of neonatal hypoglycaemia. Dr Susan Aldridge reports.

Continuous glucose monitoring (CGM) can improve outcomes for women with type 1 diabetes and their babies by improving glycaemic control during pregnancy. For instance, the CONCEPTT trial showed that the use of real-time CGM (rt-CGM) by women with type 1 diabetes during pregnancy led to a significant reduction in large-for-gestational age (LGA) infants, neonatal hypoglycaemia and neonatal intensive care admissions. Accordingly, CGM is now widely recommended for all pregnant women with type 1.

In recent years, flash or intermittently scanned (isCGM), with the FreeStyle Libre system has become increasingly popular among people with type 1 diabetes. The cost of rtCGM has prevented its widespread implementation during pregnancy. isCGM could be a cheaper alternative and it has been approved for use during pregnancy. However, it has not yet been shown to have the same benefits during pregnancy as rtCGM, as data from trials has yielded mixed results. Therefore, Verónica Perea, Hospital Universitari Mútua de Terrassa, Barcelona, and colleagues, have looked at whether the addition of isCGM to standard care improves maternal glycaemic control and pregnancy outcomes in women with type 1 diabetes on multiple daily injections (MDI). In Spain, isCGM is government-funded for these women ­– so the findings of this study would be relevant to the economics of the healthcare system here, and elsewhere. 

This was an observational study, involving 300 pregnant women with type 1 diabetes, of whom 132 were using isCGM (FreeStyle Libre or FreeStyle Libre 2). The others used conventional finger prick self-monitoring of blood glucose (SMBG). It was the largest cohort study to date evaluating the effect on isCGM in pregnancy compared with current clinical practice.

HbA1c was measured every 4 to 8 weeks during the pregnancy and one measurement per trimester was selected for the study. Information on the participants was gathered from a web-based Spanish national registry designed by the Spanish Diabetes and Pregnancy Study Group.

The primary neonatal outcome was delivery of an LGA infant and several secondary maternal and neonatal outcomes, often associated with diabetes in pregnancy, were also measured, such as severe maternal hypoglycaemia, Caesarian section and neonatal hypoglycaemia.

Glycaemic control findings

Glycaemic control during pregnancy is crucial when diabetes is involved. All participants showed a significant decrease in HbA1c from the pregestational period to the second trimester. Then there was a slight increase from the second to the third trimester. This increase was greater in the isCGM group. However, the isCGM group actually had a lower median HbA1c in the second trimester than the SMBG group.

The researchers analysed the HbA1c data according to whether women were meeting NICE or ADA recommended targets. These differ in that NICE recommends HbA1c <48 mmol/mol throughout the pregnancy, whereas ADA recommendations are tighter, with targets of HbA1c <48 mmol/mol in pre-pregnancy and the first trimester and HbA1c <42 mmol/mol for the second and third trimesters. When NICE goals were applied, there were no differences between the two groups. But with the ADA goals, significant differences did emerge in the second trimester, where 56% of the isCGM group achieved their goal, versus 42.6% of the SMBG group. And then there was a notable reduction from the second to third trimester in women in the isCGM group meeting their target – from 56% to 36.4%, compared with 42.6% to 35.7% in the SMBG group.

Pregnancy outcome findings

Initial statistical analysis showed no difference in pregnancy outcomes between the isCGM and SMBG groups. But when the analysis was adjusted for well-known confounders like age, pregestational BMI and so on, it turned out that isCGM users had a higher risk of neonatal hypoglycaemia. There were no other differences in pregnancy outcomes between the two groups. Higher glucose levels in the peripartum period can play a role in neonatal hypoglycaemia. In this study, the isCGM users had significantly lower HbA1c in the second trimester, but there were no differences between the groups in the third trimester. So it might be that there were further increases in glucose levels in the weeks or days before delivery in the isCGM group. This might account for the increased risk of neonatal hypoglycaemia, say the researchers. 

In conclusion, this study shows that the addition of isCGM to standard care for pregnant women with type 1 diabetes shows an initial improvement in glycaemic control. However, this was not sustained. And this deterioration late in pregnancy might be linked to the increased risk of neonatal hypoglycaemia with isCGM that this study also reveals. Further studies of isCGM – including the use of the new FreeStyle Libre 2 – are now urgently needed. Until more encouraging data are available, women with type 1 diabetes should be offered the use of rtCGM over isCGM during pregnancy. 

To read this paper, go to: Perea V, Picón M, Megia A, Goya M, Wägner AM, Vega B, Seguí N, Montañez MD, Vinagre I. Addition of intermittently scanned continuous glucose monitoring to standard care in a cohort of pregnant women with type 1 diabetes: effect on glycaemic control and pregnancy outcomes. Diabetologia 12 May 2022.

For more on this topic, enrol on the following EASD e-Learning courses:

Any opinions expressed in this article are the responsibility of the EASD e-Learning Programme Director, Dr Eleanor D Kennedy.

Tirzepatide is a new class of drug for management of type 2 diabetes. A meta-analysis of its glucose-lowering and weight-loss effects, presented in a recent issue of Diabetologia, will help guide treatment decisions ahead of its imminent approval for clinical use.

Glucagon-like peptide-1 (GLP-1) and glucose-dependent insulinotropic peptide (GIP), the two main incretin hormones, are released from the gut after eating and stimulate insulin secretion. The incretin effect is decreased in people with type 2 diabetes and this has led in recent years to the development of the GLP-1 receptor agonists (RAs), which mimic the incretin effect. These drugs not only lower glucose levels, they also reduce weight and improve cardiovascular outcomes.

So, what about combining GLP-1 and GIP receptor activation to boost the incretin effect even further? For the two hormones both act in synergy, and complement one another, in pancreatic beta cells.

Enter tirzepatide, a dual GLP-1 and GIP receptor agonist that has been developed for type 1 diabetes. It has a greater affinity for GIP receptors than for GLP-1 receptors and a half-life of five days, allowing for weekly injection. Early trials have suggested that tirzepatide improves markers of both beta cell function and insulin sensitivity compared with GLP-1 RAs. And the SURPASS trial has investigated the efficacy and safety of tirzepatide in comparison with placebo and other glucose-lowering medications, including GLP-1 RAs and basal insulin.

On the basis of these findings, tirzepatide has been submitted for approval to both the Food and Drug Administration (FDA) and the European Medicines Agency (EMA) for the treatment of type 2 diabetes. Thomas Karagiannis, of the Aristotle University of Thessalonki, and colleagues have set the scene for its launch later this year with a review of all the randomised controlled trials (RCTs) involving tirzepatide against placebo or other glucose-lowering drugs, including insulin. They found seven RCTs, covering 6609 participants, whose average HbA1c was 66.47 mmol/mol and mean body weight 91.5 kg.

Tirzepatide and glycaemic control

Compared with placebo, reductions in HbA1c ranged between 17.71 mmol/mol with tirzepatide 5 mg and 22.35 mmol/mol with tirzepatide 15 mg. In terms of achieving target HbA1c, all three doses (5 mg, 10 mg, 15 mg) were superior to placebo in reaching HbA1c <53 mmol/mol, £48 mol/mol or <39 mmol/mol targets.

Tirzepatide at all three doses also reduced HbA1c more than GLP-1 RAs. And more participants receiving any tirzepatide dose achieved the two lower HbA1c targets than those on GLP-1 RAs. 

When compared with basal insulin, the three tirzepatide doses were more effective in reducing HbA1c. Mean differences ranged from 7.66 mmol/mol with tirzepatide 5 mg and 12.02 mmol/mol with tirzepatide 15 mg. Tirzepatide was also more successful than basal insulin in helping participants reach all three of the HbA1c targets.

Tirzepatide and body weight

Tirzepatide, as expected, causes significant reductions in body weight. Compared with placebo, average dose-dependent weight loss was 6.31 kg (5 mg), 8.43 kg (10 mg) and 9.36 kg (15 mg). In diabetes management, especially when the goal is type 2 remission, people with diabetes are often advised to lose a percentage of their body weight. This review showed that more participants on tirzepatide had reductions of 5%, 10% or even 15% of their body weight than those on placebo.

The GLP-1 RAs are also notable for their weight loss, as well as their glucose-lowering, effect. So how does tirzepatide compare? It actually produces larger reductions in weight than the GLP-1 RAs, ranging from 1.68 kg with 5 mg to 7.16 kg with the 15 mg dose. And all tirzepatide doses were more effective than the GLP-1 RAs in achieving a body weight reduction of at least 10% and 15%.

Tirzepatide and adverse events

All glucose-lowering drugs can potentially cause hypoglycaemia. The good news is that, with tirzepatide, incidence of any kind of hypoglycaemia was the same as on placebo and was lower than with basal insulin. Severe hypoglycaemia was rare, with only 22 such events being recorded across all the trials.

Gastrointestinal events – nausea, vomiting and diarrhoea – were more common with tirzepatide than on placebo or insulin. Rates were similar to those found with GLP-1 RAs.

When it came to discontinuation of treatment due to adverse events, more patients discontinued on tirzepatide 10 mg and 15 mg compared with placebo. In comparison with the GLP-1 RAs, only those on 15 mg tirzepatide were more likely to discontinue treatment and, in comparison with basal insulin, both the 5 mg and 15 mg were associated with greater odds of discontinuing.

Finally, the incidence of serious adverse events and mortality did not differ between any of the tirzepatide doses and any comparator. There were 41 deaths among 4573 individuals receiving tirzepatide and 39 among 2151 receiving a comparator. The authors note that 19 of these deaths were related to COVID-19.

Looking forward

This meta-analysis has shown that once-weekly tirzepatide is associated with dose-dependent reductions in HbA1c that are clinically significant. These reductions also compare favourably with those achieved by once-weekly GLP-1 RAs and basal insulin. There is also a notable dose-dependent decrease in body weight, even in comparison with the GLP-1 RAs semaglutide and dulaglutide. On the down side, however, there were gastrointestinal side effects to contend with. And the 15 mg dose was associated with at least two-fold risk of discontinuation of the drug, regardless of comparator.

Tirzepatide could receive marketing approval later this year. This meta-analysis could help healthcare professionals and other stakeholders determine where this new medication fits in diabetes care, alongside the many other medications. Its potential for weight loss is likely to be of particular interest, given that even the 5 mg dose can reduce body weight and seems to be superior in this respect to even subcutaneous semaglutide. However, more head-to-head data for tirzepatide versus GLP-1 RAs on weight loss is needed. And it should be noted that an application for a label extension of semaglutide 2 mg has been submitted to the FDA and this has recently received a positive recommendation by the EMA. So, expect further developments in the options available for weight loss in type 2 diabetes. On present evidence, however, it does look as if tirzepatide could be a reasonable choice in this context. Finally, it should be noted that the ongoing SURPASS-CVOT trial should soon provide some answers on the impact of tirzepatide, compared with dulaglutide, on cardiovascular risk – another important consideration when prescribing for type 2 diabetes.

To read this paper, go to: Karagiannis T, Avgerinos I, Liakos A, Del Prato S, Matthews DR, Tsapas A, Bekiari E. Management of type 2 diabetes with the dual GIP/GLP-1 receptor agonist tirzepatide: a systematic review and meta-analysis. Diabetologia 17 May 2022.

For more on this topic, enrol on the following EASD e-Learning course:

Any opinions expressed in this article are the responsibility of the EASD e-Learning Programme Director, Dr Eleanor D Kennedy.

A new analysis of the Global Burden of Disease study, presented in a recent issue of Diabetologia, reveals the geographical variation of type 2 diabetes and how risk factors for the associated morbidity and mortality play out in different countries. Dr Susan Aldridge reports.

Economic development and improvement in healthcare have led to a decrease in type 2 diabetes-related mortality in some countries, including Scotland, Canada, the UK, Denmark, Sweden, Australia and the USA. But large developing economies like China, report an ongoing and increasing burden of type 2 diabetes, with diabetes mortality increasing from 5.3 deaths per 100,000 people in 1990 to 10.9/100,000 in 2017. In developing countries with limited resources, lack of access to diabetes treatments is an important cause of type 2 diabetes-related morbidity and mortality. One epidemiological study has suggested that between 2010 and 2030, the number with diabetes in developing countries will increase by 69%, which is more than three times the 20% increase predicted for more developed nations.

The Global Burden of Disease (GBD) study has suggested that high body mass index (BMI) is the leading risk factor in type 2 diabetes mortality and morbidity, followed by ambient particulate matter pollution and then diet. A new study from Lei Zhang of the China-Australia Joint Research Center for Infectious Diseases and colleagues, has now used GBD data to look at the differing impacts of these risk factors in individual countries according to their income levels.

The researchers had a rich source of data to work on. GBD 2019, which covers 369 diseases, including type 2 diabetes, estimated prevalence of exposure to 87 risk factors and attributable deaths, years of life lost, years lived with disability and disability-adjusted life years (DALYs) for 23 age groups in 203 countries and territories. These were grouped into 21 regions and seven super-regions. This new study looked at 13 of the GBD risk factors and their impact on type 2 diabetes in low-income countries (LICs), lower-middle-income countries (LMICs), upper-middle-income countries (UMICs) and high-income countries (HICs). They used validated formulae to work out the proportion of type 2 diabetes burden attributed to each risk factor, known as the population attributable fraction (PAF). This is a way of figuring out how important the risk factor is in type 2 diabetes morbidity and mortality.

Geographical variation of type 2 diabetes

First of all, the global number of type 2 diabetes-related deaths increased from 0.61 million in 1990 to 1.47 million in 2019, with an annual growth rate of 4.9%. Similarly, the number of DALYs attributed to type 2 diabetes increased from 29.78 million to 66.30 million over the same time period.

Behind these overall figures lies a picture of significant geographical variation in diabetes burden. The country with the highest burden – with 257.4 deaths and 6884.3 DALYs per 100,000 person-years – is Fiji. Japan has the lowest death rate at 2.0/100,000 and France the lowest DALYs at 278.2/100,000. When stratified by regions, Oceania had the highest death rate at 121/100,000 and DALYs at 3703/100,000, while the lowest death rate – 4.2/100,000 – was in the high-income Asia-Pacific region and the lowest level of DALYS, at 376/100,000, was in Eastern Europe.  And when it came to variation by national income, type 2 diabetes-related death rates were two to three times higher in low and low-to-middle income countries than they were in upper-middle and high income countries – rates were 33 and 31.2/100,000 versus 15.5 and 9.7/100,000 respectively. A similar trend was observed with respect to DALYs.

Variation in risk factor impact

Globally, high BMI is the leading risk factor for type 2 diabetes-related mortality. Using the formula, the researcher concluded that the type 2 diabetes death rate attributable to high BMI has gone up from 5.0 to 7.6/100,000 person years between 1990 and 2019. The second risk factor was ambient particulate matter pollution, associated with an increase from 1.6 to 2.6/100,000 during this time period. Finally, low levels of physical activity were linked with a small increase in death rates from 1.5 to 1.6/100,000 person-years. Thus, these three risk factors together accounted for over half (54.1%) of all type 2 diabetes deaths in 2019 globally. Similar trends were noted for DALYs.

However, the relative impact of risk factors on type 2 burden varied depending on a nation’s income. In LICs, BMI was the leading risk factor contributing to type 2 diabetes mortality, followed by household air pollution from burning solid fuel and, finally, having a diet poor in fruits. The impact on DALYs followed a similar trend. In LMICs, high BMI was also the leading contributor to type 2 diabetes mortality, but followed by ambient particulate air pollution and second-hand smoke. This pattern was, again, replicated for DALYs.

In countries with higher incomes (UMICs and HICs), high BMI remained the leading risk factor for mortality, but there were some differences in other risk factors. In UMICs, ambient particulate pollution and low physical activity were the next two risk factors, but in HICs, these were a diet high in processed meat, then low levels of physical activity. For DALYs, smoking appeared as a risk factor for both groups. It’s also notable that for HICs, the death rate attributable to high BMI has gone down from 5.3 to 4.7/100,000 from 1990 to 2019, and its contribution is considerably less than in the other income groups.

The study also analysed the link between gross domestic product (GDP) and diabetes burden. Per capita GDP was positively correlated with morbidity and mortality arising from diet high in red meat, processed meat, sugar-sweetened beverages and low physical activity. It was negatively correlated with morbidity and mortality caused by solid fuel household pollution, a diet low in fruits and smoking. This gives us an idea of how type 2 diabetes risk factors influence outcomes depending upon income. 

Targeting risk factors

This new study shows that low- and middle-income countries had the largest annual increase in type 2 diabetes morbidity and mortality and the high-income countries the smallest in the 1990 to 2019 period. BMI aside, risk factors in low- to middle-income countries are environment-related, while those in higher income countries are mostly lifestyle related.  Typically, concerns about type 2 diabetes have been focused on higher income countries. It is maybe time to shift this focus, if the gap in disease burden between rich and poor is not to widen in the future. The poorer countries have limited medical resources to serve the growing numbers with type 2. And there is often limited awareness of diabetes among the population, which leads to significant delays in diagnosis and treatment – adding to the burden of type 2-related disability and complications. Finally, the significance of pollution as a diabetes risk factor in poorer countries arises because they have limited options when it comes to energy consumption. Pollution and poverty are closely linked, with an estimated 3 billion people in low- and middle-income countries still using firewood, biomass or charcoal and traditional stoves for heating and cooking. This leaves these people exposed to indoor air pollution, which is a known cause of insulin resistance.

Meanwhile, exposure to ambient particulate air pollution accounted for 4.2 million deaths and 103 million DALYs in 2015. Again, this type of pollution mainly affects the poor, who are often neglected as countries strive for economic development. This study strongly suggests that sustainable development and pollution reduction should be prioritised to reduce the type 2 diabetes burden.

Obesity remains the leading risk factor for type 2 morbidity and mortality across all geographical regions and income groups. Historically, the increase in obesity began in high-income countries in the 1970s, followed by medium-income countries and then, more recently, by the poorer nations. These increases are likely a result of gradual economic development around the world.

However, there is a positive side – increased wealth can also mean better healthcare, including investment in type 2 diabetes treatment and prevention. Over the past few years, some high-income countries including the US, UK and The Netherlands, have reported success in type 2 diabetes remission and prevention through weight-loss programmes. Such programmes, tackling obesity, have now begun in lower income countries. For instance, in 2016, the Chinese government began its ‘Healthy China 2030’ programme in an attempt to stem the rapid growth of type 2 diabetes there. This is urgently needed, for China currently accounts for 26.2% of people living with diabetes globally.

This new study also highlights that a diet rich in processed meat is the second most important risk factor for type 2 diabetes burden in HICs. The link between processed meat consumption and type 2 has been found in other studies. As global meat prices have fallen, red and processed meat has become more readily available in middle and low-income countries too. Increased consumption could lead, in turn, to a further increase in the type 2 diabetes burden in these countries.

These new findings from the GBD study shed light on the impact risk factors have on the burden of type 2 diabetes in countries of different income levels. It’s to be hoped that they will inspire the development of strategies for action to tackle these risk factors. Reducing BMI needs to be a priority across the board. Higher income countries need to focus more on lifestyle issues and middle to lower-income countries need action on environmental pollution. And the health of the population, particularly regarding type 2 diabetes, should not be sidelined in the rush for economic development.

To read this paper, go to: Liu J, Bai R, Zhonglin C,  Cooper ME, Zimmet PZ, Zhang L. Low- and middle-income countries demonstrate rapid growth of type 2 diabetes: an analysis based on Global Burden of Disease 1990–2019 data. Diabetologia 19 May 2022.

For more on the global burden of diabetes, see the following series on Horizons:

Any opinions expressed in this article are the responsibility of the EASD e-Learning Programme Director, Dr Eleanor D Kennedy.

Insulin allergy can cause significant challenges in diabetes treatment. A new paper in Diabetologia discusses the diagnosis and management of this rare, but important, condition. Dr Susan Aldridge reports.

Insulin allergy is rare, with an estimated prevalence of 0.1 to 3%. But that amounts to 800,000 cases worldwide, so it’s a problem of clinical significance in diabetes, as it obviously complicates its treatment. Symptoms of insulin allergy generally consist of immediate or delayed skin reactions or, less commonly, systemic reactions – including life-threatening anaphylaxis. Although insulin itself is the main allergen, excipients, nickel in needles and the latex used in vial caps and cartridge plungers may also be involved.

Insulin allergy is poorly understood and there is currently no diagnostic or treatment consensus on how it should be approached. Accordingly, Agnès Sola-Gazagnes, at the Assistance Publique-Hôpitaux de Paris, and colleagues, carried out a retrospective cohort study to validate clinical criteria that will help identify those individuals who need a specialised allergology referral. They then carried out a second study to evaluate the diagnostic performance of allergology tests for insulin allergy.

Insulin allergy – or not?

The researchers recruited 52 consecutive patients who had been referred to their clinic with suspected insulin allergy between 2000 and 2010. They studied the details of the participants’ reaction to an insulin injection, using four criteria based on the clinic’s experience and guidelines for evaluating drug allergies. Based on the presence of all, some or none of these criteria, they were able to place the participants into three classes of insulin allergy – clinically likely, possible and unlikely. There were 26 participants in the clinically likely category, of whom most had local reactions, though seven did have systemic reactions – five had generalised urticaria and two had swelling of the larynx, which would be graded as severe anaphylaxis. Clinically possible insulin allergy was assigned when some, but not all, of the criteria were met and this applied to nine participants, who all reported itching at their injection site, but without a skin response that could be verified by the a physician.

Finally, the remaining 17 patients met none of the four criteria, so were classed as being unlikely to have a real insulin allergy. They tended to have skin reactions away from the injection site or to have non-specific or delayed reactions, and for these not to occur consistently with each injection.

Who is most likely to have insulin allergy?

The distribution of age, type 1 or type 2, diabetes duration, HbA1c and history of atopy were similar across the three categories. The delay between the first injection of the index insulin and allergic reaction varied but was commonly three to four months. Those in the clinically likely group developed a reaction sooner and this was judged to be the most relevant clinical feature of a true insulin allergy. The index formulations causing the allergic reactions included all major types of insulin, but insulin detemir and protamine-containing insulins featured most frequently.

The diagnoses were confirmed by intradermal reaction (IDR) and skin prick tests, using 10 different insulin formulations. The IDR test proved positive in 24 of the 26 participants in the clinically likely category, with the two who tested negative experiencing delayed responses to insulin detemir, which probably explains the negative results. Three of the nine in the possible insulin allergy participants tested positive and all of those in the unlikely insulin allergy tested negative. All the skin prick-positive patients were also IDR positive. Finally, anti-insulin IgE was measured in a subgroup of patients, with 12 out of the 15 likely insulin allergy patients testing positive.

The researchers extended the study by assessing the diagnostic performance of the IDR, skin prick and anti-insulin IgE tests with a case control study. This involved participants with clinically likely insulin allergy compared with insulin-naïve people with type 2 diabetes and non-allergic insulin-treated people with type 1 diabetes. From the findings of both studies, the researchers conclude that an IDR test alone can be used to confirm the presence of insulin allergy, and there is no need to add skin prick and anti-insulin IgE tests. The clinical likelihood criteria described above can therefore be used to effectively guide diabetologists towards an insulin allergy diagnosis, before going to allergology testing.

Management of insulin allergy

The researchers describe a stepwise management approach for people who have a positive IDR test to a particular insulin formulation (or formulations). They included 31 participants with clinically likely insulin allergy – 23 from the retrospective study and 8 from the case control study – and another 3 with possible insulin allergy. Mean follow-up duration of this management study was 4.7 years, ranging from 0.5 to 12 years. First, spontaneous resolution of insulin allergy was observed in three patients. Second, replacement of insulin detemir with an oral hypoglycaemic agent, GLP-1 receptor agonist or another insulin formulation resolved symptoms in five patients. The need for insulin was re-evaluated in another five patients and an oral hypoglycaemic agent or GLP-1 receptor agonist was substituted. Another three patients were just switched to another insulin formulation. Antihistamine treatment cleared up symptoms in four patients with immediate-type reactions. The remaining patients were switched to insulin pump treatment with either insulin aspart or insulin lispro. The researchers conclude that insulin allergy, once properly diagnosed, can be managed by a switch to oral medication, GLP-1 receptor agonists or another insulin formulation. But, eventually, insulin pump therapy will probably be necessary for most patients.

To read this paper, go to: Sola-Gazagnes A, Pecquet C, Berré S, Achenbach P, Pierson L-A, Vimoux-Buisson I, M’Bemba J, Elgrably F, Moguelet P, Boitard C, Caillat-Zucman S, Laanani M, Coste J, Larger E, Mallone R. Insulin allergy: a diagnostic and therapeutic strategy based on a retrospective cohort and a case control study. Diabetologia 4 May 2022.

Any opinions expressed in this article are the responsibility of the EASD e-Learning Programme Director, Dr Eleanor D Kennedy.