Category: Journals

A new study in Diabetologia finds that people with type 2 diabetes who are socioeconomically disadvantaged are less likely to be prescribed incretin-based therapies, including GLP-1 receptor agonists, even though they may have more to gain from such treatments. The authors suggest a number of ways of moving forward so that better value for money for society can be obtained from these new cardioprotective therapies. Dr Susan Aldridge reports. 

Low socioeconomic status has a negative impact on morbidity and mortality and is also a risk factor for type 2 diabetes via mediators including obesity, alcohol, reduced physical activity, stress, low health literacy and limited access to healthy food and exercise facilities. The result is often poor diabetes management and increased risk of cardiovascular complications.

Disparities in care, including the unequal use of incretin-based therapies, are also influenced by socioeconomic status. The incretin-based therapies are glucose-lowering drugs, including glucagon-like peptide-1 receptor agonists (GLP-1 RAs), dipeptidyl peptidase-4 (DPP-4) inhibitors and the recently developed dual glucose-dependent insulinotropic polypeptide (GIP)/GLP-1 RAs, such as tirzepatide. The ADA and EASD recommend use of agents that have demonstrated cardiovascular benefit in those individuals with type 2 diabetes with cardiovascular risk. Certain GLP-1 RAs are included among these agents and emerging data suggests that tirzepatide, too, may have cardioprotective effects.  

In a new review, Apostolos Tsapas and colleagues at Aristotle University of Thessaloniki, Greece, have summarised real-world evidence on the use of incretin-based therapies in clinical practice across the socioeconomic spectrum. They look at socioeconomic disparities in the adoption of these therapies, the possible factors driving these and how these might be addressed in the future. The study consisted of a literature search focusing on incretin-based therapy use with regard to the following aspects of socioeconomic status: area-level indexes, income, education, sociodemographic variables. 

Area-level indexes and income

One study in the US showed that individuals with type 2 diabetes and cardiovascular disease with increased area-level socioeconomic deprivation were less likely to receive GLP-1 RAs compared with those living in more privileged areas. In Australia, the Index of Relative Socioeconomic Disadvantage ranks areas according to information on income, education, employment, occupation, housing and other indicators. A study assessing the relationship between use of incretin-based therapies and the Index found that those in the most disadvantaged areas were consistently less likely to receive GLP-1 RAs, while the opposite was so for DPP-4 inhibitors.

The study also found a connection between low socioeconomic status and a reduced probability of receiving SGLT-2 inhibitors, while no such relationship was observed for metformin, sulphonylureas or insulin. And, according to a UK study, when fully accounting for various confounding factors, those belonging to the most deprived group, as identified by the Index of Multiple Deprivation, had a lower likelihood of being prescribed GLP-1 RAs.  

In another study from the US, the odds of being treated with a GLP-1 RA were greater among individuals with diabetes who had a high household income compared with those whose annual income was less than $50,000. Higher income individuals were also more likely to be on GLP-1 RAs in the All of Us Research Program, a US contemporary cohort study. Finally, in Denmark, metformin-treated patients with a high household income were more likely to initiate second-line treatment with a GLP-1 RA compared with those with a low household income. 

Education and sociodemographic factors

Education is often used as an indicator of socioeconomic status because it captures the knowledge-related assets of a person and determines their future employment, occupation and income. In the US, those with a high-school education had lower odds of receiving a GLP-1 RA prescription in comparison with those who had a postgraduate degree. Similarly, in the All of Us Research Program, a higher percentage of those who went to college were on GLP-1 RAs compared with those with less than a high-school diploma. In Denmark, the probability of initiating a GLP-1 RA was higher in those with a college education compared with lower educational levels. 

In addition, multinational data from the global DISCOVER programme suggested that those who had less than 13 years of education had lower odds of receiving a GLP-1 RA than a sulphonylurea. Similar, although less marked, associations with education level and drug utilisation were also found for DPP-4 inhibitors, but not for insulin.

Meanwhile, sociodemographic factors, such as race/ethnicity and age can also influence the uptake of newer medications. For instance, retrospective cohort data from the US suggest that, compared with White individuals with type 2 diabetes, Asian, Black and Hispanic individuals were less likely to receive GLP-1 RA therapy. In the UK, compared with White individuals, Asian and Black minorities were more likely to be prescribed metformin or sulphonylureas than GLP-1 RAs or SGLT-2 inhibitors. And in Denmark, inequalities in GLP-1 RA therapy between those with a high income and those with a low income were more pronounced in immigrants than the native Danish population. Finally, in the US, older age has been associated with a lower probability of receiving GLP-1 RAs or SGLT-2 inhibitors in those with type 2 diabetes and atherosclerotic cardiovascular disease.  

Drivers of inequalities in incretin-based therapy

The authors have identified some key mechanisms that underlie the inequalities reported above. Firstly, people may simply not be able to afford incretin-based therapies and this will, of course, vary according to the country where they live. In a system that provides universal reimbursement, such as the UK and many other European countries, access to these therapies may be more equitable because the treatments are available to more of the population, regardless of their financial status. However, access to GLP-1 RAs is not consistent across European populations. For example, in the UK, they can only be prescribed to people with type 2 diabetes who also have obesity, while other countries do not impose such restrictions. 

Where insurance coverage is not universal, as in the US, people with lower socioeconomic backgrounds may face barriers in accessing these therapies, due to higher out-of-pocket costs or limited insurance coverage. In the US, for instance, many states lack expanded Medicare coverage and one study has shown that even Medicare beneficiaries are less likely to receive GLP-1 RAs than those with private insurance. A similar trend has been uncovered in Germany, with private health insurance being a strong predictor of GLP-1 RA prescription.

The disparity in use of GLP-1 RAs according to income is currently being exacerbated by the current global shortage of semaglutide and dulaglutide, which is being driven, in part, by increasing off-label use for weight loss. So a considerable proportion of limited supply is being redirected towards well-off individuals seeking weight reduction, regardless of their diabetes status. This situation disproportionately affects those of lower socioeconomic status with type 2 diabetes and cardiovascular disease, who are totally dependent upon affordable reimbursement to access these medications. 

Secondly, there is the issue of actual access to incretin-based therapies. People living in rural or disadvantaged areas may find it hard to get to an appointment with a diabetes specialist who is well-versed in these new drugs. Finding a pharmacy that stocks them may be another challenge. Primary care physicians may not be fully aware of the cardiovascular benefits of incretin-based therapies and may be reluctant to prescribe them, even to their patients with cardiovascular disease. And, of course, most people with type 2 diabetes are treated in primary care. Research has confirmed that diabetologists and endocrinologists are more likely to prescribe GLP-1 RAs, perhaps because they are more familiar with injectables. 

Thirdly, disadvantaged groups have been consistently reported to have lower health literacy than more privileged groups in society and this has been associated with poorer health outcomes and decreased uptake of therapeutic and preventive interventions. Those with low health literacy may find it difficult to understand health information and communicate effectively with healthcare professionals. 

In some countries, they may be confused about insurance coverage options and the availability and out-of-pocket costs of GLP-1 RAs. These concerns may lead them to avoid or postpone treatment. Furthermore, those with low health literacy are less likely to be well-informed about the cardiovascular benefits of GLP-1 RAs and might not feel confident in trying to access, or get a referral, to a diabetes specialist in order to get a prescription. And during a consultation with a healthcare provider, they may find it challenging to raise their concerns about cost or the need for subcutaneous administration. Language and cultural barriers may also arise. 

Finally, on the other side of the consultation, there may be conscious or unconscious bias on the part of the healthcare provider. A recent study has shown that GLP-1 RAs are less likely to be prescribed to patients from lower socioeconomic backgrounds or minority groups, who are perceived to be less compliant to treatment and medical advice, or unable to afford the cost of these drugs. 

Increasing societal benefit from incretin-based therapy

Reflecting on the above, the authors propose some strategies for increasing the uptake of incretin-based therapy. There are consistent data from clinical trials supporting the use of incretin-based therapies, particularly GLP-1 RAs, in socioeconomically disadvantaged people. They have much to gain as they are at increased risk of cardiovascular complications in type 2 diabetes and should therefore be a priority for receiving these treatments. 

Addressing barriers to accessing incretin-based therapies is vital for increasing and widening their uptake. This might mean providing transportation to appointments for those living in rural or low-income regions and increasing the availability and retention of both primary care and specialist staff in underserved areas. Meanwhile, primary care physicians’ familiarity with GLP-1 RAs should be improved through targeted education, training and resources, given that most people with type 2 diabetes are treated in primary care. 

The authors emphasise that the availability of beneficial medications at low cost is the key to increasing their value for money from a broader societal perspective. This is especially pertinent when it comes to considering the newer, high-cost incretin-based therapies such as semaglutide or tirzepatide. 

A recent study from the US concluded that, as a first-line therapy, the cost of GLP-1 RAs needs to fall by at least 70% to be cost-effective in comparison with metformin. Another Australian study says that GLP-1 RAs are not cost-effective at current prices for either primary or secondary cardiovascular prevention – although SGLT-2 inhibitors are. Similar findings emerged from an analysis of data from 67 low- and middle-income countries. Finally, a review from high-income countries has suggested that GLP-1 RAs were not cost-effective compared with DPP-4 inhibitors, sulphonylureas or thiazolidinediones. 

These economic evaluations highlight the need for country-specific strategies to improve the cost-effectiveness of GLP-1 RAs. These would include collaborative efforts between governments and pharmaceutical companies to lower drug prices, establish product listing agreements and promote generics. The authors note that the manufacturer of liraglutide and semaglutide has recently enjoyed a significant increase in market cap, reflecting its robust financial position. Thus, there is scope for negotiating lower prices for these medications without adversely impacting company profitability or scope for research and investment.

Improving health literacy in people with type 2 diabetes can help them appreciate the importance of preventing cardiovascular complications and help them actively participate in discussion with their physicians about potentially cardioprotective drugs. Through shared decision-making, patients may be more able to accept co-payments for these drugs and overcome barriers such as the need for injection with GLP-1 RAs. This collaborative approach can enhance treatment adherence and persistence, ultimately leading to better outcomes. 

Finally, it is necessary to address physician-patient barriers. This can be a complex issue, requiring a multifaceted approach, including implicit bias training and promoting diversity within healthcare professions. The former helps healthcare professionals to become aware of any unconscious prejudice that might be affecting their decision-making and contributing to disparities in prescribing incretin-based therapies. The latter creates a more inclusive healthcare environment, which may help promote better understanding of those from disadvantaged groups.  

In conclusion

The authors have highlighted disparities in the uptake of incretin-based therapies, particularly GLP-1 RAs, in people with type 2 diabetes according to socioeconomic status. An essential first step in addressing this problem is advocating for a reduction in the price of GLP-1 RAs, which would improve their value for money for society at large. This approach would complement other measures, such as increasing accessibility, improving health literacy and overcoming physician-patient barriers. 

However, much of the evidence informing this paper comes from the US and findings may not be generalisable to other settings, so additional research on uptake disparities and context-specific strategies for improvement is needed. Collaborative efforts to implement these strategies will boost the societal outcomes of the incretin-based therapies and improve health outcomes for those with type 2 diabetes.

To read this paper, go to: Karagiannis T, Bekiari E, Tsapas A. Socioeconomic aspects of incretin-based therapy. Diabetologia 12 July 2023. https://doi.org/10.1007/s00125-023-05962-z

To learn more, enrol on the EASD e-Learning course ‘GLP-1 receptor agonists’.

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

One in five people with type 2 diabetes have a normal or low body weight, with reduced muscle mass relative to their fat mass. A new study reported in Diabetologia shows that they have more to gain from strength training than weight loss when it comes to glycaemic control. Dr Susan Aldridge reports. 

While the majority of people with type 2 diabetes are overweight or obese, around 20% have a healthy weight with a BMI of 25 or less. This is known as normal-weight type 2 diabetes – it is now recognised as being a particular phenotype and it’s more common among Asians and older people. Normal-weight type 2 diabetes is associated with sarcopenia or loss of muscle mass, and research has suggested that this feature mediates the elevated mortality risk seen in this phenotype compared with diabetes with overweight.

Exercise is always recommended for people with type 2 diabetes. Guidelines are similar to those for the general population: three to five days per week of moderate-to-vigorous aerobic activity to reach a minimum duration of 150 min per week, plus two to three sessions of strength training. Trials comparing the impact of aerobic versus strength exercise on HbA1c have mostly been carried out in people with type 2 diabetes who also have overweight or obesity. For instance, the Diabetes Aerobic and Resistance Exercise (DARE) study and the Health Benefits of Aerobic and Resistance Training in individuals with type 2 diabetes (HART-D) study both found a combination of aerobic and strength training to be superior to either modality on its own in lowering HbA1c.

Individuals with obesity have both increased fat mass and increased lean muscle mass, while those with normal-weight type 2 diabetes are more likely to have a different body composition – namely, sarcopenia, especially related to their fat mass, which is known as relative sarcopenia. This suggests that the most effective exercise training for those with normal-weight type 2 diabetes may not be the same as for those who have overweight and obesity. That is why Yukari Kobayashi at Stanford University and colleagues set up the Strength Training Regimen for Normal Weight Diabetics (STRONG-D) study, which looks at the effects of strength training alone, aerobic training alone and a combination of the two upon glycaemic control in normal-weight people with diabetes. They hypothesised that these individuals might respond better to strength training than aerobic training, given their phenotype. The study looked at changes in body composition and muscle strength from these interventions and how these impacted HbA1c.

Focus on strength training

The 186 participants, of whom 83% were Asian, were assigned to either strength training only (ST), aerobic training (AER) or a combination of the two (COMB), which they did for three days a week for nine months. Strength training consisted of two sets of upper-body exercises (bench press, seated row, shoulder press and pull-down), three sets of leg exercises (leg press, extension and flexion) and abdominal crunches and back extensions. They worked up, increasing weight, to eight to 12 repetitions in a set. 

The aerobic group worked on a treadmill or exercise bike to 50-80% of their metabolic equivalent of task, which is energy expenditure related to their weight. Combination was two strength-training sessions and a slightly reduced amount of aerobic training. The primary outcome was change in HbA1c and secondary outcomes were changes in body composition and muscle strength at nine months. 

Mean HbA1c was 59.6 mmol/mol at the start of the trial and 131 participants actually completed it with data for analysis. At the end, there was a significant mean decrease in HbA1c in the ST group of 0.44%, compared with non-significant decreases of 0.35% in the COMB group and 0.24% in the AER group. Therefore, strength training alone was better than aerobic training or a combination of the two in reducing HbA1c levels in normal-weight individuals and combination training had an intermediate effect. Strength training also increased appendicular (arms and legs) lean mass relative to fat mass and was an independent predictor of a reduction in HbA1c. 

This was the first clinical trial of exercise in normal-weight individuals but there was no significant difference in the AER or COMB group. Of course, the findings need to be confirmed in further, larger studies, but there is no reason not to apply a recommendation of strength training immediately to people with normal-weight diabetes, the authors say.

In the STRONG-D study, only the ST group showed a significant reduction in HbA1c, which suggests a potentially unique benefit of strength training in normal-weight individuals with type 2 diabetes. In comparison with the strength-training groups in the HART-D and DARE studies, the ST group achieved a higher absolute mean reduction in HbA1c. The participants in STRONG-D had a lower fat mass and much lower lean mass than the participants in the other two studies. Given that 80% of the insulin-mediated glucose uptake occurs in skeletal muscle or lean mass, it may be important to look at increasing lean mass for improving glycaemic control in this population. 

Focus on body composition

An important finding of STRONG-D is that body composition – increase in lean mass, loss of fat mass – was independently associated with a reduction in HbA1c, while an increase in lean mass or decrease in fat mass alone was not. This adds to the growing body of evidence that estimates of muscle mass adjusted for fat mass show stronger associations with metabolic abnormalities than lean mass alone. 

Strength training led to increased muscle mass relative to decreased fat mass and it is this that seems to be more beneficial for lowering HbA1c in individuals with normal-weight HbA1c. In contrast, individuals with overweight or obesity can lower HbA1c by lowering fat mass. At present, there isn’t enough data to support body composition as a central target for exercise training in type 2 diabetes. However, these new findings, along with previous studies that have shown a relationship between body composition and cardiovascular mortality, show the benefits of strength training in the normal-weight diabetes population. 

Furthermore, weight loss is well established as being associated with a reduction in HbA1c in people with overweight or obesity and type 2 diabetes. In the STRONG-D study, significant weight loss was found only in the AER group and there was no relationship between weight loss and reduction in HbA1c. Thus, the most effective exercise regimen for overweight and obese individuals with type 2 diabetes may not necessarily be applicable to normal-weight individuals with type 2 diabetes. 

The findings of STRONG-D make an important contribution to exercise recommendations for lean individuals with type 2 diabetes. They could also inform personalised exercise recommendations for different diabetes phenotypes. In the current clinical guidelines for people with type 2 diabetes, there are no recommended strength-training regimens, so the authors used the strength-training programme from the HART-D study. 

In conclusion, the STRONG-D study shows that strength training alone was effective and superior to aerobic training alone for reducing HbA1c levels in individuals with normal-weight type 2 diabetes. Individuals with normal-weight type 2 diabetes present with relative sarcopenia and achieving increased muscle mass relative to decreased fat mass via strength training plays an important role in glycaemic control in this population. The findings of this study could help refine physical activity recommendations in type 2 diabetes by weight status.  

To read this paper, go to: Kobayashi Y, Long J, Dan S, Johannsen NM, Talamoa R, Raghuram A, Chung S, Kent K, Basina M, Lamendola C, Haddad F, Leonard Mb, Church TS, Palaniappan L. Strength training is more effective than aerobic exercise for improving glycaemic control and body composition in people with normal-weight type 2 diabetes: a randomised controlled trial.

Diabetologia 26 July 2023. https://doi.org/10.1007/s00125-023-05958-9

To learn more, enrol on the EASD e-Learning course ‘Management of hyperglycaemia in type 2 diabetes’.

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

The presence of severe mental illness can worsen cardiovascular morbidity and mortality suggests a study based on a nationwide type 2 diabetes cohort, reported in Diabetes Care. This high-risk group therefore merits special attention when it comes to prevention and management of heart disease. Dr Susan Aldridge reports.

People who have severe mental illness (SMI), including schizophrenia, bipolar disorder and major depression, have a 10 to 20 years shorter life expectancy than the general population. This premature mortality is largely due to a higher risk of physical disease, particularly cardiovascular disease (CVD), where diabetes is a major risk factor. Having SMI is also associated with a two to three-fold higher risk of type 2 diabetes – a risk that may even be on the increase.  

While we already know that depression in people with diabetes is associated with increased risk of CVD and cardiac death, other complications and all-cause mortality, fewer studies have focused on severe depression, schizophrenia and bipolar disorder. In the few existing studies, SMI is consistently associated with an increased risk of mortality among people with diabetes, but findings on associations between SMI and macrovascular and microvascular complications have been inconsistent. 

In a new study, Caroline Jackson of the University of Edinburgh and on behalf of the Scottish Diabetes Research Network Epidemiology Group, looked at a large, nationally representative diabetes cohort to determine the association between SMI and clinical outcomes. Their aim was to determine the independent effects of schizophrenia, bipolar disorder and major depression on the risk of major CVD events, CVD-specific mortality and all-cause mortality in people with type 2 diabetes. 

The study drew upon data from the Scottish Diabetes Research Network National Diabetes Dataset (SDRN-NDS), which covers 99% of people with diabetes in Scotland. It includes information on type of diabetes, sociodemographics, routine diabetes care, including retinopathy screening, and linked acute and psychiatric hospital records and death records. The study population comprised 259,875 adults diagnosed with type 2 diabetes in Scotland between 2004 and 2018 whose data could be linked with hospital and death records. 

The primary outcomes were major CVD events – myocardial infarction (MI) or stroke – in the whole cohort and also in the subgroup without a history of CVD, CVD-specific mortality and all-cause mortality. Secondary outcomes were related to other diabetes complications, consisting of retinopathy, renal replacement therapy and lower-limb amputation.

Characteristics of people with SMI and type 2 diabetes

The SDRN-NDS hospital record data revealed that 2,621 (1%) of this cohort had a diagnosis of schizophrenia, 1,211 (0.5%) had bipolar disorder and 7,903 (3%) had depression. The cohort was mainly of White ethnicity and there was a higher prevalence of type 2 diabetes among those from more deprived areas. This was even more striking among those with SMI. For instance, more than one-third of those with schizophrenia came from the most deprived fifth of areas in Scotland. 

Diabetes was diagnosed at a younger mean age in those with a history of schizophrenia (52.1 years), bipolar disorder (57.5 years) or depression (58.9 years), compared with those without a history of SMI (60.8 years). 

Furthermore, history of prior CVD, comorbidity and high cholesterol at the time of diabetes diagnosis were also more common among those with depression and bipolar disorder compared with those without SMI. This wasn’t seen for those with schizophrenia, maybe reflecting their younger age at diabetes diagnosis. Smoking, history of alcohol use disorder and overweight or obesity were also more common among those with SMI.

The impact of SMI

The researchers carried out a statistical analysis that adjusted for all the factors that could affect the findings – sociodemographics, HbA1c, hypertension, smoking and so on. This fully adjusted model helped clarify the link between the presence of SMI and health outcomes. 

During a mean of 6.9 years of follow-up, there were nearly 25,000 major CVD events. All three SMIs were associated with an increased risk of having a major CVD event. The hazard ratios were 1.07, 1.37 and 1.22, for schizophrenia, bipolar disorder and depression, respectively, for the fully adjusted model. The association was similar, whether or not the individual had a history of CVD. 

There were also 51,029 deaths occurring during a mean follow-up of 7.1 years. Deaths from CVD were higher among those with SMI than those without, with hazard ratios of 2.38 for schizophrenia, 1.70 for bipolar disorder and 1.84 for depression. This was for a model adjusted for sociodemographic factors – the hazard ratios were attenuated only slightly when the fully adjusted model was applied, so there is a persistent higher risk of CVD mortality in all groups with SMI and type 2 diabetes. Similar increased risk also applied to all-cause mortality. 

When it came to the secondary outcomes, numbers with lower limb amputation and renal replacement therapy were low – at around 0.6% and 0.2%, respectively – in those with and without a history of SMI. Referable retinopathy occurred in around 5% of both groups. Numbers for these three complications were too low to make comparisons between those with and without SMI in this study. 

Previous studies of SMI and macrovascular complications among people with diabetes report conflicting findings. For instance, a study from South Korea found a similar excess risk of heart attack and stroke among those with diabetes and schizophrenia, bipolar disorder or depression. A Taiwanese study found an association between clinical depression with diabetes and increased risk of acute coronary syndrome and stoke. And these new findings also echo those of a Danish study that found that SMI, as a composite exposure, was associated with higher CVD risk – although that was more broadly defined than in the current study. In contrast, however, there have been two studies that actually reported a lower risk of CVD events among people with schizophrenia or major depression. 

Only two previous studies have reported an association between SMI and CV-specific mortality in diabetes, with similar findings. There has also been a meta-analysis of studies examining depression of any severity and the risk and the risk of CV mortality in people with diabetes – again with similar findings. And the observed excess all-cause mortality among people with diabetes and SMI compared with no mental illness is consistent with previous literature. This new study adds to scarce data on bipolar disorder, which has been less studied in this context in comparison with schizophrenia and major depression. 

Underlying reasons

The mechanisms behind the increased CVD morbidity and mortality among those with diabetes and SMI shown by this new study are complicated and poorly understood. We already know that shared risk factors for poor physical and mental health include low socioeconomic status, adverse childhood experiences and lifestyle. In this study, the higher prevalence of smoking, overweight and obesity and comorbidities was made evident through statistical analysis that adjusted for these factors and showed an attenuation of the risk. Meanwhile, a study from Denmark of people with type 2 diabetes showed that excess mortality among those with depression was largely explained through smoking, physical activity and comorbidities. 

There is also emerging evidence that brain insulin resistance might be part of the pathophysiology of schizophrenia and bipolar disorder. This might go some way to explaining poorer diabetes outcomes in those with SMI. 

Inequalities in care for physical disease may also explain the poorer outcomes among those with SMI and diabetes. A recent study from Denmark reports lower rates of diabetes monitoring and achievement of HbA1c and cholesterol targets in those with SMI, compared with those without. However, disparities in care do not explain the findings in this new study – in Scotland, receipt of diabetes care processes is actually similar or better in those with SMI than for those without. 

It may be that this does not translate into optimal treatment of those with CVD risk factors or established CVD. Previous studies of populations with and without diabetes have uncovered suboptimal CVD risk management in those with SMI. The excess risk of CV death in people with diabetes and SMI found in this new study may reflect more severe cardiovascular events and potential differences in cardiac care, both in the acute phase and subsequently. For instance, previous work from these authors has shown that patients with SMI were less likely to receive coronary revascularisation after MI, were less likely to survive for 30 days post-MI and were more likely to have a further vascular event than those without mental illness. 

Cardiac and metabolic adverse effects of some antipsychotic medications may also play a role in poorer outcomes. The impact of antipsychotic and antidepressant drugs upon CV and mortality outcomes in people with diabetes is an underinvestigated topic. 

This new study shines a light on an area where previous studies are scarce or contradictory and addresses a number of limitations and gaps in the literature. It draws on data from a nationally representative cohort of people with diabetes with and without SMI. This large study population and long follow-up allowed the researchers to investigate individual SMIs, analyse specific CVD outcomes and obtain reliable, precise effect estimates for outcomes. The richness of the diabetes register allowed associations to be adjusted for key lifestyle factors, which hasn’t always been done in other studies. Finally, the study also adds to scarce data on associations between SMI and CVD-specific mortality in people with diabetes and on all-cause mortality in those with bipolar disorder specifically. 

In conclusion, among people with new-onset type 2 diabetes, those with a prior history of SMI have a markedly higher risk of major CVD events, CVD-specific mortality and all-cause mortality than people with no mental illness. Some of this excess risk is due to modifiable risk factors, including smoking, alcohol misuse and obesity, highlighting the need for effective lifestyle modification in people with SMI. However, there are other emerging mechanisms, including possible shared pathophysiology between SMI and diabetes, which require further investigation. 

Another future avenue of research should be the role of psychotropic medication use and receipt of optimal cardiac care in primary and secondary care settings. Meanwhile, effective prevention and management of cardiovascular risk factors is needed in this high-risk group to improve clinical outcomes. 

To read this paper, go to: Fleetwood KJ, Wild SH, Licence KAM, Mercer SW, Smith DJ, Jackson CA on behalf of the Scottish Diabetes Research Network Epidemiology Group. Diabetes Care 2023;46:1363–1371. https://doi.org/10.2337/dc23-0177

To learn more, enrol on the EASD e-Learning course ‘Cardiovascular health and diabetes’.

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

A new study, based on type 1 diabetes audit data and reported in Diabetes Care, shows that failure to engage with services, hyperglycaemia and diabetic ketoacidosis peak in late adolescence. Therefore, innovative approaches to optimising glycaemia during the transition from paediatric to adult care are needed to avoid later-life complications. Dr Susan Aldridge reports. 

In young people who have type 1 diabetes, hyperglycaemia increases during adolescence and continues to do so as they transition to adult services. Data from the T1D Exchange Clinical Network confirm this and their 2016-2018 report shows that glycaemic control in those aged 15 to 18 years has deteriorated further, despite increased use of diabetes technology.  

It is generally assumed that it is the transition from paediatric to adult services at the age of 16 to 18 years, with changes in continuity and support, which leads to hyperglycaemia at this stage of life. There are also personal psychological issues and the influence of growth-related hormonal change on insulin requirements to consider. Whatever the cause, even these few years of hyperglycaemia can have long-term consequences, as seen from the follow-up of the Diabetes Control and Complications Trial.

In a new study, Naomi Holman at Imperial College London and colleagues looked at data from diabetes audits in England and Wales to investigate the age-related changes in HbA1c measurement, HbA1c levels and hospital admissions for diabetic ketoacidosis (DKA) in children and young people.

Exploring the age-related trajectory of hyperglycaemia

The National Diabetes Audit (NDA) for England and Wales collects data on people of all ages with a diabetes diagnosis from primary care records and adult specialist diabetes services. Each audit includes data for a 15-month period from 1 January to 31 March in the following year. Meanwhile, the National Paediatric Diabetes Audit (NPDA) collects data on children and young people from paediatric diabetes services in England and Wales for a 12-month time period from 1 April to 31 March in the following year. 

The researchers identified sequential cohorts of people with type 1 diabetes aged between five and 30 years from the 2017/2018, 2018/2019 and 2019/2020 data collections of the NDA and NPDA, from which they created a pooled cohort of 93,125 individuals. Then, for each year of age, the latest HbA1c measurement was identified and recorded, along with hospital admissions for DKA. 

The existence of an HbA1c record was used as a measure of clinic attendance and engagement with health services. The proportion of records lacking an HbA1c measurement was low (around 5%) up to age 16 and then it rose rapidly, peaking at 22.3% at age 21 for men and 17.3% at age 19 to 20 years for women. It then fell gradually to 17.9% for men and 13.1% for women at the age of 30. This reflects a peak in lack of engagement with services around the ages of 19 to 21, which never falls back to the level achieved in childhood.  

As has been previously observed, HbA1c rises with age, peaking in young adulthood, then falling again. At the age of nine, median HbA1c was 60 mmol/mol for males and 61 mmol/mol for females, peaking at age 19 at 72 mmol/mol for young men and 74 mmol/mol for young women. By the age of 39, it had fallen to 68 mmol/mol for men and 66 mmol/mol for women. Between the ages of 16 and 22 years, median HbA1c was consistently higher for women than men. 

When it came to glycaemic control, those aged 20 had the lowest proportion reaching a target of 58 mmol/mol or lower. And the greatest proportion of those with a very high HbA1c (more than 86 mmol/mol) was among young men aged 17 (31.4%) and young women aged 18 (35.8%). These age-related patterns persisted after adjustment for ethnicity, social deprivation and duration of diabetes. 

Finally, annual prevalence of one or more hospital admissions for DKA rose from 2% for boys and 1.4% for girls at the age of six, to a peak of 7.9% at age 19 for young men and 12.7% at age 18 for young women. After this, prevalence fell and, at age 30, it was 4.3% for men and 5.4% for women. Each year from the age of nine, the proportion of females with one or more annual hospitalisations for DKA was significantly higher than in males.

A significant drop in attendance

This analysis of over 93,000 individuals in England and Wales highlights the changes that occur when children with type 1 diabetes pass through adolescence to early adulthood. We’ve learned that almost all children have annual records for HbA1c – only 5% do not – reflecting faithful attendance at clinic. When they transition to adult services between the ages of 16 and 20, this proportion drops to around 80%. Median HbA1c, along with hospitalisation for DKA, starts to rise earlier but peaks around the time of transition. It then decreases again, but never goes back to childhood levels. 

So the HbA1c upward trend starts before, and continues after, the transition period. This hyperglycaemia pattern reflects that found in the SWEET project, which compiled data on 66,418 individuals with type 1 diabetes from 22 centres across 19 countries. It found that, from 2016 to 2018, HbA1c rose steadily from the age of six to 18 years. The scale in the rise of HbA1c from age nine to 18 years was 12 mmol/mol for males and 11 mmol/mol for females, and is similar to the 11 mmol/mol increase in HbA1c reported in another study for a combined cohort of individuals living in Germany or Austria and the US. 

The sudden increase at age 17 to 19 years in those lacking annual HbA1c records suggests a dramatic reduction in attendance at clinic at the time of changing provision from paediatric to adult. This age is also a time of increasing psychological and social pressures – leaving home for university, for instance. Adolescence is a time of profound psychological change for all young people and the more so if they have to manage a chronic condition like diabetes. The need for continual self-care is relentless and can create a considerable burden. Both very high HbA1c levels and DKA have been associated with insulin omission and psychological stress. An additional personal and clinical challenge is the effects of the surge in puberty-associated growth hormones. This demands physiologically matched adjustments of insulin, which will only add to the burden. 

While this study was not designed to identify the underlying reasons for lack of engagement, the authors suggest that it is likely due to a combination of personal and health-service factors. Whatever the reasons, this lack of engagement in late adolescence is concerning – research shows that people with type 1 diabetes for whom routine care processes have not been recorded are at increased risk of future morbidity and mortality. This is particularly so for those with the highest HbA1c levels and repeated admissions for DKA. 

This study found greater adverse changes among young women with type 1 diabetes. They were less likely to reach a glycaemic target of 58 mmol/mol and more likely to have an HbA1c of more than 86 mmol/mol. And, between the ages of 15 and 25, they were more likely to have at least one episode of DKA, compared with men of the same age. One reason for this inequality might be the higher prevalence of disordered eating among women. Psychological and societal pressures, including approaches to body image and peer pressure, may play out differently between men and women in this age group. 

The authors point out that this new study may even underestimate problematic hyperglycaemia among children and young people. The data comes from routine health records, which means that it is limited to those individuals who are actually engaging with health services. It therefore misses those who are potentially at greatest risk. The peak age for highest HbA1c coincides with the peak age for lack of records. One-third of the cohort had HbA1c of 86 mmol/mol or more at a time when another fifth had not even had their HbA1c recorded. While we can’t know for sure, it is likely that non-attenders have high HbA1c. So the scale of the increase in HbA1c in the late teens might be even greater than that recorded in this study. 

In conclusion, these findings point to the need for novel approaches to the support of children and young adults with type 1 diabetes. The aim should be to flatten the rise in HbA1c during the teenage years and reduce the peak in DKA admissions that occurs around the same time. It is clear from this study that improved, innovative, well-resourced and age-appropriate service designs oriented to deliver optimal diabetes care over the course of adolescence to young adulthood are warranted.

To read this paper, go to Holman N, Woch E, Dayan C, Warner J, Robinson H, Young B, Elliott J. National trends in hyperglycaemia and diabetic ketoacidosis in children, adolescents and young adults with type 1 diabetes: a challenge due to age of stage of development or is new thinking about service provision needed? Diabetes Care 2023; 46(7): 1404–1408. https://doi.org/10.2337/dc23-0180

To learn more about how to manage patients after the transition into adult care, enrol on the EASD e-Learning course ‘Management of type 1 diabetes in adults’.

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

Real-world data is changing the landscape of research into diabetes outcomes. The opportunities and challenges offered by this new approach are discussed in a recent article in Diabetes Care. Dr Susan Aldridge reports.

Despite advances in research, technology and care, diabetes remains an increasing burden around the world. Addressing this burden, with its associated costs, calls for new efforts across the spectrum of research disciplines to improve diabetes care. 

In recent years, there has been a rapid growth in systems for accumulating real-world data (RWD) and using it to identify real-world evidence (RWE) outside of research settings. Collection and analysis of big data are increasing the options for evidence-based guidance of diabetes management and prevention. Although these new data are not collected for research, they do have the potential to enrich the body of health-related information. This RWD is being applied to a number of disciplines, including clinical epidemiology, health services research, population surveillance and implementation research.  

These developments have the potential to improve diabetes prevention and treatment by increasing the range of populations, settings, interventions, outcomes and clinical settings that can be explored. However, it must be borne in mind that the level of evidence that may be derived from RWD is ultimately a function of the quality of that data and rigour of study design and analysis. 

In a new report, Edward Gregg of RCSI University of Medicine and Health Sciences, Dublin, and colleagues elsewhere, look at the current landscape and applications of RWD in clinical effectiveness and population health research in diabetes. They then summarise opportunities and best practises in the conduct, reporting and dissemination of information derived from RWD to optimise its value and limit its drawbacks.

Applying RWD

Randomised controlled trials (RCTs) have long been the gold standard behind treatments, health services and policies. They take a linear path, starting with a research question and study design, then data collection and analysis. Real-world evidence studies may also start with a question, but can expand the scope of research objectives because they include a broader range of variables. 

There are four ways in which RWD can complement traditional approaches. First, it seeks to understand the effectiveness of evidence-based practices as they are necessarily modified for delivery outside the research setting. Second, it extends evaluation to population subgroups that may be under-represented in trials. Third, RWD also aims to assess the effectiveness of interventions and policies where conventional experimentation is too impractical, costly or unethical. Finally, it permits the examination of outcomes that are beyond the scope of traditional survey, trial, cohort or other measurement methods. 

Clinical diabetes research

While RCTs remain the gold standard for determining the efficacy of diabetes treatments and establishing guidelines, they do have a number of drawbacks. They study effects under ideal healthcare practice circumstances and often with highly selected populations, thereby resulting in findings that are not really generalisable to routine clinical care. They are also costly, lengthy, expose participants to potential risk and are usually designed to answer a narrow hypothesis. Higher risk older people with multiple comorbidities and polypharmacy are often excluded, although these will be the very recipients of the medications on trial when they reach the mainstream of diabetes care. For instance, in the PROactive, ADVANCE and EMPA-REG OUTCOME trials (of pioglitazone, preterax and diamicron, and empagliflozin, respectively), the clinical characteristics of the study population would have represented only 3.5%, 15.7% and 35% of the real-world population, according to one analysis. And the protocols of RCTs, which standardise treatment dosage and timing, may not be representative of management in a broader range of settings. Nor are issues relating to concurrent illness and medications and other aspects of care addressed. 

These limitations call for RWE to complement the findings of RCTs. Current healthcare systems produce a large amount of longitudinal, patient-level, electronic data that assess exposures of interest and associated health outcomes in clinical practice. These RWD can create large databases that can be used for research to improve diabetes care. 

Most published applications of RWD have compared the effectiveness of pharmacological interventions. For instance, recent cardiovascular outcome trials of SGLT-2 inhibitors have demonstrated improvements in cardiorenal outcomes. These findings have been confirmed and extended by non-interventional studies based on RWD through the inclusion of unrestricted populations as treated in actual clinical settings. These studies have resulted in effectiveness estimates for previously unstudied populations in the context of clinically varying treatments, dosages, comorbidities and dilutions of effect that can occur due to suboptimal adherence. In this setting, RWD may offer more useful evidence to complement RCT findings for those involved in planning a treatment strategy. In another example, comparative effectiveness analyses based on RWD provided support for the broader relevance of the CAROLINA study, which compared the efficacy and cardiovascular safety of a DPP-4 inhibitor and a sulphonylurea in type 2 diabetes.  

Thanks to the very large populations represented by RWD collected during routine care, RWE is also being used increasingly by regulatory agencies to assess safety of medical products, particularly with respect to rare adverse drug events that RCTs are not powered to examine. For example, diabetic ketoacidosis among users of SGLT-2 inhibitors was identified as a safety concern post-marketing by RWD studies. 

Finally, RWD studies can provide insights into the natural history of diabetes and these have extended our understanding of the risks associated with hypoglycaemia beyond what was found in RCTs. For example, the Hypoglycaemia Assessment Tool (HAT) study has shown rates of overall, nocturnal and severe hypoglycaemia that were much higher than seen in RCTs. Further exploration of these findings could allow earlier and more accurate identification of those at risk of hypoglycaemia.  

Surveillance and monitoring

Surveillance has many applications in diabetes care – including gathering data on risk factors, incidence and prevalence, complications and death among those with the condition. It can also help monitor levels of care, disparities in care access and inform policy, resource allocation and clinical decision-making. Traditionally, surveillance has relied on surveys and registries, but now we can leverage various sources of RWE, such as electronic health records in well-defined healthcare delivery systems databases (for example, Veterans Administration and Kaiser Permanente in the US). These data, collected during the routine provision of care to large populations, provide passive surveillance that can complement the traditional active surveillance carried out by the long-running National Health and Nutrition Examination Survey and other well-established studies in the US and elsewhere. 

However, surveillance based on RWD requires a good understanding of the data source. For instance, surveillance based on electronic health records works well in acute settings, such as the emergency room or hospital, but less so for outcomes that may not even come to medical attention. Surveillance of severe hypoglycaemia based only on emergency room visits or hospitalisation has been estimated to miss around 95% of episodes because they were cared for at home or by ambulance staff. And, despite the expansion of the range of RWD sources for diabetes surveillance, they remain limited in their ability to assess behaviours and patient-centred outcomes.

Much of the precedence and proof of concept for national health systems-based registries comes from early work in Scotland, the UK and Scandinavia. Here, and elsewhere, the presence of single-payer systems has facilitated the creation of comprehensive, linked registries covering primary care, lab, pharmacy, hospitalisation and mortality data. These data systems are regularly used to track trends in risk factors, care and outcomes. They also provide a basis for the study of effectiveness of interventions. This is in stark contrast to the situation in low- and middle-income countries, and in high-income countries without a single-payer system, where national or population-level systems data are rare – so they cannot yet benefit from the application of RWE. 

Challenges in RWD use

Since RWD are not usually collected for research, their use in generating evidence does present some challenges. For instance, even posing the research question can be difficult because the population being studied is less under the control of the investigator than in an RCT. Then the choice of study design will depend upon the study question and the availability of data. In comparative effectiveness studies and health services research, the goal is often to emulate an RCT and often uses a new-user, active-comparator cohort design, which focuses on new users of alternative treatments with similar indicators. This study design is analogous to an RCT, but without the baseline randomisation. Other study designs – for instance, using propensity score to ‘stand in’ for randomisation – are also being explored. 

When it comes to population monitoring, compared with cohort studies based on primary data collection, RWD-derived cohorts are usually more representative and have larger numbers of participants. However, drawbacks include missing data, non-standardised exposures and lack of linked behavioural, genetic or patient-centred outcomes. RWD tend to be dominated still by diagnostic and service data related to the processing of payments and generally lack patient-reported data on behaviour, risk, function and quality of life that could dramatically improve its utility.  

Furthermore, individual-level social determinants of health, including income, education, social care and voluntary-sector interventions, can impact surveillance outcomes, but these may not be routinely collected in RWD sources. This means there are large gaps in understanding the drivers of, and progress in, achieving health equality. 

Data harmonisation across different sources of RWD can be challenging. For example, in some settings, race and ethnicity are merged whereas elsewhere they may be treated as separate variables. Validated instruments are needed to facilitate collection of patient-reported outcomes in healthcare settings to improve future surveillance efforts. 

Meanwhile, new sources of RWD are emerging. For instance, some healthcare settings are initiating remote glucose monitoring programmes and cloud-based continuous glucose monitoring downloaded to the electronic health record. This opens up the possibility of new models of care and new research opportunities – but, again, work will be needed on data harmonisation.

The final challenge lies in the reporting of RWD to ensure transparency and credibility. The Professional Society for Health Economics and Outcomes Research (ISPOR) and the International Society for Pharmacoepidemiology have created a task force to make recommendations for good practices that would increase confidence in RWE. In their report, they look at the planning, implementation and dissemination of hypotheses evaluating treatment effectiveness in RWD studies. 

In conclusion

RCTs address a narrow hypothesis in a highly selected population, so we can’t be sure that the intervention under study will be effective and safe in practice in the wider population. RWE can help provide additional complementary evidence of efficacy and safety. In diabetes, RWE has become increasingly relevant and important in recent years in various areas, including clinical effectiveness, long-term surveillance and monitoring and regulatory affairs.  

However, there are concerns around real-world studies, including poor data quality and inconsistent methodology, which affect the reliability of findings. Standardising the structure, registration and reporting of such research is essential to improve the quality of RWE. There are other challenges, such as facilitating broader data linkage and access, as well as a need for analytical methods to keep up with newer RWD sources, such as patient apps. Notwithstanding, RWE has the potential to revolutionise patient care in diabetes as its acceptance and credibility increase.  

To read this paper, go to: Gregg EW, Patomo E, Karter AJ, Mehta R, Huang ES, White M, Patel CJ, McElvaine AT, Cefalu WT, Selby J, Riddle MC, Khunti K. Use of real-world data in population science to improve the prevention and care of diabetes-related outcomes. Diabetes Care 46(7):1316–1326. https://doi.org/10.2337/dc22-1438

To learn more, enrol on the EASD e-Learning course ‘Real-world evidence’. 

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

A clinical trial reported in Diabetologia compares once-weekly insulin icodec with once-daily insulin glargine U100 in type 2 diabetes with respect to hypoglycaemia. Dr Susan Aldridge reports.

Once-weekly insulins promise to simplify insulin therapy and could thereby address both therapeutic inertia and improve treatment adherence. However, there are potential concerns about hypoglycaemia with weekly administration, particularly if a dose is miscalculated because that would potentially expose the individual to a much larger amount of insulin than usual. 

In type 2 diabetes, longer duration of diabetes and increased time since initiation of insulin are both associated with diminishing hypoglycaemia awareness and blunted counterregulation. This may be why the risk of hypoglycaemia increases as type 2 diabetes progresses. It is therefore important that any new insulin, like the weekly insulins, be investigated to ensure that any hypoglycaemia it induces does elicit a robust symptomatic and counterregulatory response. 

Insulin icodec is a basal insulin under development for once-weekly administration. This is made possible because, on injection, insulin icodec binds strongly and reversibly to albumin, thereby creating a depot from which insulin is slowly and continuously released with a half-life that is suitable for once-weekly dosing. The action of insulin icodec is similar to that of native human insulin. Also, Phase 2 trials have shown similar reductions in HbA1c and fasting plasma glucose in type 2 diabetes as insulin glargine, and similar rates of level 2 and level 3 hypoglycaemia. 

However, given the novelty of insulin icodec, its characteristics with respect to hypoglycaemia require further investigation. Accordingly, Thomas Pieber at the Medical University of Graz, Austria, and colleagues elsewhere, carried out a randomised crossover trial of insulin icodec versus once-daily insulin glargine U100 in a group of participants with type 2 diabetes. They looked at the frequency of hypoglycaemia, time to hypoglycaemia and recovery time during a short period of exposure to double and triple doses of icodec and double and triple doses of glargine. In a subgroup analysis, they also investigated symptomatic and hormonal counterregulatory responses between these induced-hypoglycaemia experiments.  

Icodec versus glargine – hypoglycaemia profiles

The trial participants, aged between 18 and 72, were all on basal insulin with or without other glucose-lowering drugs and had no clinically significant diabetes complications or history of cardiovascular disease. They were considered reasonably representative of the type 2 diabetes population. As this was a crossover trial, they received insulin glargine or insulin icodec for either 11 days or six weeks, then swapped over after a washout period. 

While receiving each of the insulins, the hypoglycaemia-induction ‘challenges’ were carried out during short periods in which they received the double and triple doses. First, euglycaemia was achieved by allowing a glucose infusion to run to a glucose level of 5.5 mmol/l, at which point it was stopped and the glucose allowed to fall to no less than 2.5 mmol/l where it was maintained for 15 minutes. Then the euglycaemia was restored by the glucose infusion. During the hypoglycaemia window, various measurements and tests were carried out.

Hypoglycaemia was induced in 43 and 42 participants after a double dose of icodec and glargine U100, respectively, and in 38 and 40 participants after triple doses. Clinically significant hypoglycaemia, defined as glucose less than 3.0 mmol/l or the presence of symptoms, occurred in comparable numbers after exposure to the double and triple doses of both insulins. 

There were no significant treatment differences in the time to hypoglycaemia, which varied from 2.9 to 4.5 hours for the double doses and 2.2 to 2.4 hours for the triple doses. And time to recovery from hypoglycaemia, via glucose infusion, took less than 30 minutes for all treatments. Outside these hypoglycaemia-induction periods, there were no severe hypoglycaemia episodes and the rate of clinically significant hypoglycaemia episodes was comparable for icodec and glargine U100 at around three episodes per participant-year of exposure. 

Counterregulatory responses

After the double doses of icodec and glargine U100, 20 and 19 participants experienced clinically significant hypoglycaemia, as did 20 and 29 participants after the triple doses. These individuals made up the subgroup for investigation of the physiological responses to hypoglycaemia. There was significant overlap in that 16 participants after double dose and 17 after triple dose experienced hypoglycaemia after both icodec and glargine U100.  

Concentrations of all five counterregulatory hormones – glucagon, adrenaline, noradrenaline, cortisol, growth hormone – appeared to increase from baseline during hypoglycaemia induction for both insulin products following double dose and triple dose. Following a triple dose of insulin, the hormone response was greater with icodec for adrenaline and cortisol. However, there were no statistically significant differences in other counterregulatory hormone concentrations between icodec and glargine after the double or triple doses. 

Study backs once-weekly icodec on safety

So, in summary, this study shows that the risk of hypoglycaemia was comparable for icodec and glargine U100 following experiments with double and triple doses. The duration of decline in plasma glucose towards clinically significant hypoglycaemia was similar for both insulins, as was the recovery time of around 30 minutes. There was also a robust endocrine and symptomatic response among participants with clinically significant hypoglycaemia.

Phase 2 clinical trials of up to 26 weeks in people with type 2 diabetes show low rates of level 2 and level 3 hypoglycaemia for weekly icodec that are comparable to once-daily glargine U100. The results of this new study support these findings. They also suggest that even if there were a substantial mismatch between the insulin dose administered and that required, there would still not be an increased risk of hypoglycaemia with icodec compared with once-daily glargine U100. 

Avoiding hypoglycaemia is important for people with diabetes. If an episode does occur, it is important to be able to rapidly correct the fall in plasma glucose. Given the longer dosing interval and longer duration of action of once-weekly insulin, prolonged hypoglycaemia, slow recovery and recurrent events might be a big concern. However, results from this study and from Phase 2 trials are reassuring. An analysis of continuous glucose monitor (CGM) data from two Phase 2 trials showed similar duration of hypoglycaemic episodes with icodec versus glargine U100 in type 2 diabetes. And, in the current trial, recovery from hypoglycaemia via constant intravenous glucose infusion took less than 30 minutes for all treatments. Furthermore, no severe hypoglycaemia episodes were observed following the hypoglycaemia-induction episode. 

The present findings apply where an unintentionally high insulin dose is administered by either forgetting a dose has been taken and taking another one or by miscalculation. They may also apply where the usual dose is too high for current needs. This could occur during and after exercise, during intercurrent illness or during hospitalisation, if fasting is required before tests or operations. 

Double and triple doses of once-weekly icodec used in this study are actually equivalent to 14 and 21 times the regular daily dose of insulin. Despite this vast amount, when the next weekly dose was omitted in this trial, the risk of hypoglycaemia was not higher in comparison with double or triple doses of once-daily glargine U100 followed by omission of the next daily dose. And recovery from hypoglycaemia did not require more glucose for icodec versus glargine U100 after a triple dose, and only slightly more after the double dose. 

Previous studies have looked into the symptomatic and endocrine responses to hypoglycaemia for insulin detemir, insulin degludec and glargine U100 in healthy individuals and in type 1 diabetes. There have also been studies investigating the response to hypoglycaemia induced by human soluble insulin under clamp conditions in people with type 2 diabetes, but this is the first study to look at the physiological response to hypoglycaemia induced by a basal insulin analogue in type 2 diabetes. It was reassuring to discover that the physiological response to hypoglycaemia was as large for icodec as for the widely used basal insulin analogue glargine U100.

A strength of this trial was the recruitment of people with type 2 diabetes. They may have the most to gain from a weekly insulin as they – and maybe their healthcare provider – may be reluctant to intensify their treatment by starting on daily insulin injections.  

In conclusion, then, it is reassuring to learn that double and triple doses of once-weekly icodec do not lead to increased risk of hypoglycaemia compared with once-daily glargine U100. Furthermore, the time taken to develop and recover from hypoglycaemia is comparable for the two insulins. The study also suggests that hypoglycaemia induced by icodec is accompanied by symptomatic and counterregulatory responses at least as robust as those seen for glargine U100. This new study therefore provides further reassurance about the safety of once-weekly icodec, bringing it one step closer to mainstream diabetes care.

To read this study, go to: Pieber TR, Arfelt KN, Cailleteau R, Hart M, Kar S, Mursic I, Svehlikova M, Urschitz M, Haahr H. Hypoglycaemia frequency and physiological response after double or triple doses of once-weekly insulin icodec vs once-daily insulin glargine U100 in type 2 diabetes: a randomised crossover trial. Diabetologia 13 June 2023. https://doi.org/10.1007/s00125-023-05921-8

To learn more, enrol on the EASD e-Learning course ‘Hypoglycaemia’.

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

A new review in Diabetologia looks at the very earliest stages of type 1 diabetes, focusing on the beta cell injury and the inflammatory immune response that set the scene for an autoimmune attack. This recent research will inform strategies for primary prevention, taking us ever closer to a world without type 1 diabetes. Dr Susan Aldridge reports.

The countdown to type 1 diabetes begins with a pre-symptomatic autoimmune attack on the islets, which can be diagnosed through the detection of one or more islet autoantibodies. The presence of more than one of these antibodies is known as stage 1 type 1 diabetes and is associated with a high probability of developing clinical diabetes over the following years. 

There are now treatments, such as teplizumab, which can be administered at stage 1 or later to prevent or delay the onset of overt diabetes. However, the more we can learn about the pre-symptomatic stages, the more likely it is that such treatments can be optimised, thereby reducing the burden of type 1 diabetes in the population. Professor Anette-Gabriele Ziegler of the Institute of Diabetes Research, Helmholtz Munich in Germany, has reviewed when, how and why islet autoantibodies arise and how our knowledge to date can be used to plan future strategies to target islet autoimmunity. This review is based on her 2022 EASD-Novo Nordisk Foundation Diabetes Prize for Excellence presentation. 

Professor Ziegler draws upon three longitudinal birth cohorts for her findings. The BABYDIAB/BABYDIET study began in 1989 and has over 30 years of prospective follow-up. It included 2441 children with a first-degree relative with type 1 diabetes. The Environmental Determinants of Diabetes in the Young (TEDDY) study, which began in 2004, is an international birth cohort of 8676 children with increased risk for type 1 diabetes defined by their human leukocyte antigen (HLA) status. Finally, the Primary Oral Insulin Trial (POInT), which began recruitment in 2018, has enrolled 1050 children with a high polygenic risk score for type 1 diabetes from Belgium, Germany, Poland, Sweden and the UK. All of these studies are following children for the development of islet autoantibodies and clinical type 1 diabetes.

Islet autoimmunity – when?

Both the BABYDIAB and TEDDY studies showed that incidence of islet autoimmunity peaks between the age of one to 1.5 years, and this is so for both children with a first-degree relative with type 1 diabetes and children from the general population. These data suggest that there is a ‘fertile’ period very early in life in which there is enhanced susceptibility to autoimmune attack on beta cell targets. 

The author and her colleagues did a modelling study that showed that the risk of developing islet autoantibodies decreases exponentially with age. This has practical relevance for screening and advising families. For instance, a baby born to a father with type 1 diabetes has an estimated risk of 7% of developing islet autoantibodies by the age of six. However, if they are still negative at this age, the remaining risk of developing them over the next six years falls to just 1%. The age of peak incidence and exponential decay with age also have important implications for the aetiology and pathogenesis of type 1 diabetes, as the genes and/or environmental factors influencing islet autoimmunity must exert maximum impact at, or prior to, the age of peak incidence.  

The role of genetics

A type 1 diabetes-associated genotype is the first important component of the fertile environment. Before BABYDIAB, it wasn’t clear whether these genes were responsible for the development of islet autoimmunity or progression to type 1 diabetes. It now appears that genetic susceptibility is important for the initiation of autoimmunity and less so for progression. Both HLA and non-HLA genes are involved in developing islet autoantibodies. The prevalence of multiple islet autoantibodies in children from the general population is around 0.31%, but children with specific HLA genotypes have an average risk of 5% of developing multiple islet antibodies by the age of six. 

Adding non-HLA genes to polygenic risk scores can further stratify the risk, thereby identifying newborns in the population with a 10% risk of developing multiple autoantibodies by the age of six. Genetic risk changes with age, suggesting that a substantial component of this risk operates during the fertile period. Most of these genes are associated with immune or cellular responses to infection. Some are associated with altered microbiome compositions or increased birthweight. This is interesting, given that both the microbiome and birthweight are known to be associated with development of islet autoimmunity and type 1 diabetes. 

Focus on the fertile period

The genes that predispose for type 1 diabetes are also implicated in other autoimmune diseases, including coeliac disease, pernicious anaemia and thyroid disease. The incidence peaks of autoantibodies differs for each condition – two years of age for coeliac disease and puberty and adolescence for thyroid disease. This does suggest that the affected organ is relevant and its activity, function and maturity may change during its fertile period. The author’s team wanted to know why beta cell autoimmunity occurs so early in life. Therefore, they measured random preprandial non-fasting glucose levels between four months and 3.5 years in children who were enrolled in POInT. 

They found substantial changes in glucose concentrations – falling to a nadir at one to 1.5 years of age and increasing thereafter. This not only coincides with the peak incidence of islet autoimmunity, but also follows the adiposity peak that is typically seen at eight to nine months of age, suggesting growth pressure on islets during this susceptible period. Therefore, they propose that age one year might be a period of increased islet cell activity, increased beta cell stress and greater vulnerability to insults. 

Respiratory infection is one of these insults, being most frequent during the first two years of life. Indeed, children in the BABYDIAB/BABYDIET studies had an average of 3.7 infection episodes during the peak period of islet autoimmunity. This is not unexpected as the protection afforded by maternal antibodies gradually declines as the child’s own immune system develops. It is thought that some viruses directly infect beta cells, while others indirectly cause beta cell damage and immune activation via systemic inflammation.  

A range of studies, taken together, build a picture where there are deviations from a healthy immune response, with propensity to a T helper type 1 (Th) response to insulin and marked inflammation present before the onset of islet autoimmunity. 

The role of viruses

Previous research has linked viral infections in the first year of life with type 1 diabetes, particularly among children who have prolonged or multiple respiratory infections. There is evidence to suggest that childhood infections occurring shortly before the onset of autoimmunity play a role in promoting autoreactivity towards islet cells. 

Many attempts have been made to identify the viruses responsible for the increased risk of autoimmunity and the candidates with the most convincing evidence to date are the enteroviruses, particularly coxsackie B virus. For instance, a large sequencing study of stool samples from 700 children in the TEDDY study suggests that prolonged shedding of enterovirus B may be involved in islet autoimmunity. However, this was observed in only 11.8% of those with autoimmunity – as opposed to only 6.5% of children without islet autoantibodies – so enterovirus B is clearly only part of the story.  

Research has also suggested that other viruses may be involved in islet autoimmunity. For instance, an increased incidence of type 1 diabetes was observed during the COVID-19 pandemic. The receptor for the SARS-CoV-2 virus is expressed on pancreatic ductal, alpha and beta cells, so it is plausible that it may trigger an autoimmune response. Finally, rotavirus and cytomegalovirus have also been implicated in islet autoimmunity. A decrease in type 1 diabetes incidence has been noted with the introduction of rotavirus vaccination in some, but not all, studies. In summary, the available data suggests that viruses, especially those that can infect islet cells, are a co-factor, with genetics, for the development of islet autoimmunity. 

Implications for prevention

So what does this recent research mean for prevention of type 1 diabetes? The Global Platform for the Prevention of Autoimmune Diabetes (GPPAD) was set up in 2017 to identify infants at higher genetic risk of developing type 1 diabetes and to carry out primary prevention trials. The GPPAD uses a polygenic risk score to screen newborns for risk of type 1 diabetes in five European countries and over 400,000 newborns have joined screening to date. Two randomised controlled trials are underway. POInT, as mentioned above, aims to induce tolerance to insulin by oral insulin therapy and is intended to target the Th 1 propensity to insulin in children with high genetic susceptibility. Results are expected in 2025. 

The second trial, Supplementation with Bifidobacterium infantis for Mitigation of Type 1 diabetes Autoimmunity (SINT1A), aims to reduce inflammation and infection by promoting a healthy microbiome through supplementation with B.infantis from the age of six weeks to 12 months. The study has enrolled 600 infants out of a target of 1150, with results expected in 2027. More studies are being planned and one goal of these would be to protect the beta cell from infection or stress to prevent the initiation of autoimmunity.

In summary, there are three factors that contribute to the development of beta cell autoimmunity. There is increased beta cell activity, exposure to infection and a heightened immune response with a propensity for Th 1 immunity. Beta cell injury and activation of an inflammatory immune response therefore sets the scene for autoimmunity. A deeper understanding of these early events in the countdown to type 1 diabetes will aid development of effective primary prevention strategies, bringing us closer to Professor Ziegler’s vision of a world without type 1 diabetes. 

To read this paper, go to: Ziegler A-G. The countdown to type 1 diabetes: when, how and why does the clock start? Diabetologia 26 May 2023. https://doi.org/10.1007/s00125-023-05927-2

To learn more, enrol on the EASD e-Learning course ‘Diagnosis of type 1 diabetes’.

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

A new study reported in Diabetologia shows that the association between prediabetes and dementia can be explained by the development of clinical diabetes. Preventing or delaying the progression of prediabetes to diabetes could help reduce the burden of dementia. Dr Susan Aldridge reports.

Prediabetes carries a high risk of progression to diabetes and is also independently associated with other clinical outcomes. Previous research has shown that prediabetes is a risk factor for cognitive decline and dementia, but few studies have looked at the impact of the transition from prediabetes to clinical diabetes upon the subsequent risk of dementia. It’s therefore not clear whether the intervening development of diabetes accounts for the excess risk of dementia in people with prediabetes. 

The risk of diabetes complications also depends on the age at which someone is diagnosed. Earlier onset diabetes generally means a more severe course of diabetes with a higher risk of complications compared with later onset diabetes. The age of onset is thus crucial in evaluating the risk of long-term outcomes such as dementia.

Elizabeth Selvin of Johns Hopkins University and colleagues elsewhere in the US used data from a large population-based cohort to look at the link between prediabetes and dementia. Their study examines how far the intervening development of diabetes can explain this association. They also looked at the link between dementia risk and incident diabetes according to age at diabetes onset.  

The Atherosclerosis Risk in Communities (ARIC) study is a prospective study of 15,792 participants aged 45 to 64 from four US counties within different states, recruited between 1987 and 1989. Baseline for the dementia study was the second visit, in which cognitive function and HbA1c were measured. There were 11,656 participants in the final analytical sample. Prediabetes was defined as HbA1c between 39 and 46 mmol/mol. Incident diabetes was self-reported or diabetes medication-use reported during a regular study visit or phone call. Dementia was diagnosed through hospital reports or cognitive function assessments during study visits. 

Prediabetes, diabetes – then dementia

It was found that 20% of the study group had prediabetes and during nearly 16 years of follow-up, 3143 participants developed diabetes. Those with prediabetes were, as expected, more likely to be diagnosed with diabetes than those without – 44.6% versus 22.5%. A total of 2247 participants developed dementia over a follow-up time of nearly 25 years. 

Among those with prediabetes, the cumulative incidence of dementia was 16.6% higher in those who went on to develop diabetes compared with those who did not. The cumulative incidence of dementia in those with prediabetes was 15%, and 10% among those without prediabetes by the age of 80 years and 63% and 53%, respectively, by the age of 90. Analysis showed that after adjusting for incident diabetes, the association of prediabetes and dementia was no longer significant. In other words, developing diabetes is the key transitional link between prediabetes and dementia. 

The cumulative incidence of dementia was highest among those who developed diabetes at an earlier age. The strength of the association between incident diabetes and dementia decreased with older age at diabetes onset. Those who were diagnosed with diabetes younger than 60 years of age were at the highest risk of dementia. Those diagnosed at age 80 or older did not have a significantly increased risk of dementia compared with those who did not go on to develop diabetes.

Therefore, prediabetes is associated with dementia, but this can be explained by its progression to clinical diabetes. This study shines a light on the link between prediabetes and dementia, showing that prediabetes in itself is not a risk factor in the absence of a subsequent diagnosis of diabetes.

Prevention is key

In the US, up to 96 million adults have prediabetes – 38% of the adult population. We know that structured lifestyle intervention programmes, such as the National Diabetes Prevention Program, can effectively prevent diabetes progression. However, fewer than 5% of those with prediabetes are receiving referrals to such programmes from their healthcare providers. And more than 80% of adults are unaware that they even have prediabetes. These new findings make an urgent case for pushing prediabetes higher up the public health agenda in the US and elsewhere. 

There are a few other studies in line with this new evidence. For instance, the Whitehall II study has shown that every five-year earlier onset of diabetes is significantly associated with a higher risk of dementia. And the Swedish Twin Registry study reported a greater risk of dementia among people whose age of diabetes onset was less than 65. 

Possible mechanisms by which prediabetes and diabetes lead to dementia include acute and chronic hyperglycaemia, glucose toxicity, insulin resistance and microvascular dysfunction of the central nervous system. Glucose toxicity and microvascular dysfunction are associated with increased inflammatory and oxidative stress, which lead to increased blood-brain permeability.

A combination of all these mechanisms has been proposed to explain the link between diabetes and both vascular and Alzheimer’s dementia. However, further studies are needed to clarify the underlying pathophysiology of prediabetes followed by diabetes and dementia.  

In conclusion, this new study underlines the importance of early detection of prediabetes and engagement in prevention of its progression to clinical diabetes as an approach to the prevention of dementia in later life.  

To read this study, go to: Hu J, Fang M, Pike JR, Lutsey PL, Richey Sharrett A, Wagenknecht LE, Hughes TM, Seegmiller JC, Gottesman RF, Mosley TH, Coresh J, Selvin E. et al. Prediabetes, intervening diabetes and subsequent risk of dementia: the Atherosclerosis Risk in Communities (ARIC) study. Diabetologia 24 May 2023. https://doi.org/10.1007/s00125-023-05930-7

To learn more about conditions that are associated with diabetes, enrol on the EASD e-Learning course ‘Multimorbidity and diabetes’.

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

Research on the impact of physical activity on health outcomes is hampered by its reliance on participants’ self-reporting of their daily exercise, which is subjective. A new study, reported in Diabetes Care, quantifies physical activity with data from a wearable tracker and shows that there is a linear relationship between daily energy expenditure on physical activity and reduced risk of type 2 diabetes, particularly if that activity is more intense. Dr Susan Aldridge reports. 

‘Move more’ has long been part of standard advice for prevention of type 2 diabetes as there is a well-established inverse relationship between physical activity and incidence of the condition. However, most research to date has relied on self-reporting of physical activity, which is subject to bias and difficult to quantify. This makes it hard to be precise in public health messaging when it comes to how much physical activity is needed to make a significant impact on type 2 diabetes incidence. 

The ‘gold standard’ method for measuring physical activity energy expenditure (PAEE) involves the use of calorimetry with radiolabelled water (the ‘stable isotope’ method), which is impractical and expensive for use in large population studies. However, the use of wearables, such as the accelerometers popularised by Fitbits and similar devices, now offers a viable alternative to quantifying dose-response associations of PAEE with health outcomes. 

Few studies have taken this new approach to looking at the impact of PAEE on type 2 diabetes risk. Those that exist have been small and none has validated measurement methods against the gold standard. Accordingly, Søren Brage and Nick Wareham of the Institute of Metabolic Science at Cambridge University have carried out a new study to look at the association between accelerometer-derived PAEE and incident type 2 diabetes in a large cohort of middle-aged adults without diabetes at the start of the study. 

Quantifying daily physical activity

The study population was drawn from the UK Biobank, a prospective health study of over half a million adults recruited between 2006 and 2010. At five years after recruitment, a subsample of 90,096 individuals was invited to wear a wrist accelerometer for a seven-day period. 

Data was summarised according to proportions of daily time spent at different movement intensity levels, from which the researchers calculated PAEE in units of kilojoules per kilogram per day. They also validated their measurements against the stable isotope method in a sub-group of 97 participants. In addition, they worked out the fraction of PAEE from moderate-to-vigorous physical activity (MVPA), as this is an important component when it comes to health outcomes, including type 2 diabetes. Meanwhile, the Hospital Episodes Statistics database, or participant self-report, was used to record incident diabetes up to November 2020. There were 2,018 cases of type 2 diabetes occurring during this time period. 

The researchers also investigated a number of factors that could potentially influence the relationship between PAEE and incident type 2 diabetes. These included sex, age, ethnicity, body mass index (BMI) and genetic predisposition to type 2 diabetes. They also looked at grip strength, cardiovascular fitness, cardiovascular disease and cancer status. This level of detail would allow physical activity advice to be tailored more precisely to the individual. 

Take an extra brisk walk

This was a large, prospective study with objective measurement of PAEE and it revealed a linear inverse relationship between PAEE and the risk of type 2 diabetes, both with and without adjustment for BMI. Furthermore, there was no attenuation in the relationship, even at higher PAEE levels. 

Quantifying PAEE showed that each 5 kJ/kg/day was associated with a 19% lower risk of type 2 diabetes and 11% lower when BMI was adjusted for. This is equivalent to an additional 20-minute brisk walk per day. The findings suggest that the benefits of ‘moving more’ are constant, whatever an individual’s initial level of physical activity. Put simply, ‘some is good, but more is better’. 

The strength of the association between PAEE and type 2 diabetes did differ by sex, BMI and genetic susceptibility to obesity, but the linear inverse relationship persisted among all the subgroups investigated (save for those of non-White ethnicity, where there were too few participants to reach a significant conclusion).

Further analysis showed that intensity of physical activity also had an impact. Accumulating the same PAEE through a higher intensity of exercise was associated with lower risk of type 2 diabetes than accumulating it through a lower intensity activity. This is in line with the authors’ previous research on all-cause mortality and cardiovascular disease. It highlights an important health message – health benefits can be achieved through a variety of combinations of volume and intensity (often put as ‘whatever you enjoy, whatever is sustainable for you’). However, if possible, go for more intense activity, as it delivers more. 

Previous research has shown that higher intensity activities may impact the risk of type 2 diabetes through metabolic adaptations, while lower intensity activities are mediated through changes in BMI. Higher intensity work means greater reliance on carbohydrate oxidation which, in turn, could increase the expression and activity of proteins involved in glucose metabolism and insulin signalling. And perhaps greater stimulation of cardiovascular-related pathways leads to improved cardiovascular fitness which, in itself, lowers the risk of type 2 diabetes. 

So, in summary, there is a strong inverse relationship between accelerometer-derived PAEE and incident type 2 diabetes in a large sample of middle-aged adults. A difference in PAEE equivalent to an extra 20-minute brisk walk per day was associated with 19% lower odds of developing type 2 diabetes. These findings support physical activity for diabetes prevention in the whole population, given that the association persisted across all subgroups, and engaging in more MVPA was associated with additional benefit. 

Therefore, activity intensity, over and above its contribution to PAEE, is particularly important for avoiding incident type 2 diabetes. These new findings add weight to health messaging encouraging physical activity for prevention of type 2 diabetes and should be taken on board by all involved in the prevention agenda.

To read this paper, go to: Strain T, Dempsey PC, Wijndaele K, Sharp SJ, Kerrison N, Gonzales TI, Li C, Wheeler E, Lagenberg C, Brage S, Wareham N. Quantifying the relationship between physical activity energy expenditure and incident type 2 diabetes: a prospective cohort study of device-measured activity in 90,096 adults. Diabetes Care 2023; 46(6):1145–1155. https://doi.org/10.2337/dc22-1467

To learn more, enrol on the EASD e-Learning course ‘Lifestyle intervention’ with ‘Module 2: Promoting physical activity for people with diabetes’ launching soon.

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 Diabetologia, looks at the glycaemic impact of different forms of exercise performed after a meal in people with type 1 diabetes. The findings should help inform more detailed guidelines as well as identifying some crucial research gaps, so that those living with type 1 diabetes can enjoy the benefits of physical activity while minimising the risks. Dr Susan Aldridge reports. 

Physical activity can be challenging for people with type 1 diabetes because of concerns around managing blood glucose levels, with exercise-induced hypoglycaemia being a particular concern. Glycaemic responses to exercise depend upon whether it is performed in a fasted or fed state, how much insulin is on board and the timing, type and intensity of the physical activity being undertaken. This level of detail does not yet appear in current guidelines. If it did, people with type 1 diabetes could enjoy more flexibility in scheduling their exercise sessions and the type of activity undertaken, which would make them more likely to enjoy its benefits, rather than avoiding it for fear of adverse consequences on blood glucose.   

Accordingly, Simon Helleputte, University of Ghent, Belgium, Jane Yardley, University of Alberta, Edmonton, Canada, and colleagues elsewhere, have summarised the current state of research into the glycaemic impact of different types of postprandial exercise (taken within two hours of a meal, with insulin administration) in people with type 1 diabetes. They reviewed 20 clinical trials on the acute (during exercise), sub-acute (within two hours of exercise) and late (more than two hours and up to 24 hours after exercise) glycaemic effects of postprandial exercise. 

Findings were organised according to four types of exercise – walking, continuous moderate intensity, continuous high intensity and interval exercise. The researchers also noted whether guidelines on reducing insulin before exercise to avoid hypoglycaemia were followed in the reviewed studies and the impact it had. They did not find any studies on resistance exercise, which is an obvious gap in the research as this is often recommended for overall fitness. 

Reductions in blood glucose across the board

The review found that blood glucose decreased during all modalities of postprandial exercise, seemingly regardless of pre-exercise bolus insulin reduction, but depending on exercise duration and intensity. Walking, outside or on a treadmill, was defined as low-intensity aerobic exercise. There were only two studies in this category and both showed that a short walk – between 15 and 30 minutes – shortly after a meal lowers blood glucose during and for two hours after exercise, thus avoiding meal-related hyperglycaemia. 

Continuous moderate-intensity exercise was defined as exercise at less than 70% of maximal aerobic capacity. The eight studies in this category showed decreases in blood glucose ranging from 2.3 mmol/l to 5 mmol/l. The larger decreases were found with longer and more intense bouts of activity. There were also many reports of premature termination of exercise because of low blood glucose, particularly when no insulin-reduction strategy was used.

Seven studies looked at the impact of high-intensity exercise, defined as exercise between 70% and 80% of maximal aerobic capacity. When carried out for between 45 and 60 minutes, large decreases in blood glucose, ranging from 3 mmol/l to 8 mmol/l were seen. Hypoglycaemia was infrequent, presumably because insulin-reduction strategies were used in most of these studies. The risk of late nocturnal hypoglycaemia persisted, sometimes even when participants were protected by a post-exercise meal. 

Interval exercise consisted of either intermittent high-intensity exercise (IHE) or high-intensity interval training (HIIT). IHE is defined as a combination of continuous, moderate-intensity exercise interspersed with very short high-intensity sprints at regular time intervals. HIIT is brief, intermittent periods of vigorous, near-maximal capacity exercise, interspersed with low-intensity recovery periods of similar duration. There were five studies – three on IHE and two on HIIT. All led to declines in blood glucose during exercise between 1.9 mmol/l and 3.9 mmol/l and, notably, none used an insulin-reduction strategy.  

So exercise is a useful blood-glucose management tool in type 1 diabetes, the authors say, with the largest effects seen in high-intensity exercise and the lowest with HIIT, but the exact decline depending on intensity and duration of exercise. However, physical exercise does carry a risk of hypoglycaemia, which can be mitigated by reductions in the insulin taken to cover the pre-exercise meal. This creates higher blood-glucose levels at the start of exercise, protecting against hypoglycaemia, but similar declines in blood glucose are seen, regardless of the insulin-reduction strategy used. 

These findings suggest that non-insulin mediated uptake of glucose by muscle is an important mechanism for driving blood-glucose decreases during exercise and it is perhaps even more important than the amount of circulating insulin. Another important point is that pre-exercise insulin reductions should not come at the expense of significant hyperglycaemia before, during or after exercise because this, over time, could increase the risk of diabetes complications. Blood glucose before exercise and the precise timing of postprandial exercise are crucial in maintaining the balance between hypoglycaemia and hyperglycaemia, and further research is now needed to look at this aspect of physical activity in type 1 diabetes.  

Future directions

This review revealed some gaps in the research on postprandial exercise in type 1 diabetes, besides getting the balance between hypoglycaemia and hyperglycaemia right, as mentioned above. Individual characteristics should be taken into account in future research, as there was considerable heterogeneity among the participants in the review studies. For instance, blood-glucose declines and risks of hypoglycaemia might vary according to an individual’s physical fitness or gender. 

Pre-exercise meal composition is another important factor. It may be that higher protein or fibre content could be protective against hypoglycaemia. In the review studies, there was great variation in the meals consumed – some being high carbohydrate and others being a more standard composition. Some did not even report the detailed composition of the meal beyond its carbohydrate content, so it is difficult to draw firm conclusions on the impact of the pre-exercise meal composition on glucose excursions thereafter. If we knew the optimal pre-exercise meal composition, it would be easy for a person with type 1 diabetes to incorporate into their exercise plan. 

The studies also had an overall lack of detailed information on the late effects of postprandial exercise, although there was evidence of nocturnal hypoglycaemia and the authors point to the value of a post-exercise snack or meal. However, continuous glucose monitoring (CGM) was not widely available when most of the review studies were carried out. With the improved current availability, future research should be able to describe the risks of post-exercise hypoglycaemia in more detail and advise people with type 1 diabetes accordingly. There will also hopefully be opportunities for research on postprandial exercise in hybrid closed-loop users, who have options to decrease their circulating insulin before, during and after exercise, simplifying glucose management around physical activity.  

So, in conclusion, this review provides new insights into the glycaemic effects of postprandial exercise to improve glucose management. Postprandial exercise results in a consistent decline in blood glucose, whatever the exercise and regardless of insulin-reduction strategy. The authors recommend substantially reducing the prandial insulin dose at the pre-exercise meal to achieve higher blood glucose before, during and after exercise, thereby preventing exercise-induced hypoglycaemia. The magnitude of this insulin reduction should be proportional to exercise duration and intensity, but pre-meal blood-glucose levels and timing of exercise should also be considered to avoid pronounced hyperglycaemia. With these strategies, people with type 1 diabetes should be able to exercise safely and enjoy its benefits, without compromising their glycaemic management.

To read this paper, go to: Helleputte S, Yardley JE, Scott SN, Stautmans J, Jansseune L, Marlier J, De Backer T, Lapauw B, Calders P. Effect of postprandial exercise on blood glucose levels in adults with type 1 diabetes: a review. Diabetologia 4 April 2023. https://doi.org/10.1007/s00125-023-05910-x

To learn more, listen to Professors Miles Fisher and Mike Riddell discuss how technology is making a difference for athletes with diabetes in ‘The long and the short of it’.

To learn more, enrol on the EASD e-Learning course ‘Lifestyle intervention’, with ‘Module 2: Promoting physical activity for people with diabetes’ launching soon.

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