Addressing the prediabetes conundrum
Prediabetes is a risk factor for type 2 diabetes, diabetes complications and mortality. It is highly prevalent and a new discussion paper in Diabetologia looks at different approaches for reducing the potential health burden that prediabetes poses to both the individual and society at large. Dr Susan Aldridge reports.
It is more than 20 years since the American Diabetes Association (ADA) replaced the terms ‘impaired glucose tolerance’ (IGT) and ‘impaired fasting glucose’ (IFG) with ‘prediabetes’ in their standards of care. The term prediabetes remains controversial. Supporters say that its use enables the realisation of preventative potential at the individual level, while those against say that labelling all those with intermediate hyperglycaemia (IHG) as having a ‘pre-disease’ medicalises a large proportion of the population. There is also a lack of consensus on how we define prediabetes and how to approach those individuals who have it.
The prevalence of prediabetes is high and those who do have it are at increased risk of developing type 2 diabetes, along with its complications and related conditions, compared with those who have normoglycaemia. But there is considerable heterogeneity in individual risk. This poses a dilemma – how to find the right balance between undertreatment of those most at risk and overtreatment, with medicalisation and unnecessary stigma.
Opinion on how to approach prediabetes varies. Some want to keep the current definition, while others argue for including more markers of risk in order to identify those most at risk. And still others would like to abandon the prediabetes concept entirely. It’s also been suggested that the diagnostic threshold of diabetes be lowered to include the prediabetic range. Martin Blond, of the Steno Diabetes Center in Copenhagen and colleagues elsewhere, shine a light on the prediabetes conundrum with a brief overview and suggestions for how to deal with it going forward.
Defining prediabetes
Diabetes is a multisystem, multifactorial condition that, for practical and historical reasons, is diagnosed only on blood glucose or HbA1c levels. Prediabetes is defined as the presence of IHG in the form of IGT, IFG or elevated HbA1c. Although there is no clear threshold for progression to type 2 diabetes, specific but varying cut-off points are used to define prediabetes.
The term IGT was introduced to cover the glucose range between normal glucose tolerance and diabetes, while IFG was introduced at a later date and was defined so that prevalence of IGT and IFG were similar in the Paris Prospective Study, a long-term investigation of cardiovascular disease.
The lower cut-off point for IFG in this study was 6.1 mmol/l, a value still used by the World Health Organization. The ADA lowered this cut-off point to 5.6 mmol/l in 2003, which increased the prevalence of IFG dramatically. In 2008, an expert committee with members from the ADA, EASD and the International Diabetes Federation (IDF) concluded that HbA1c is a reliable measure of hyperglycaemia and one that is associated with long-term complications, so it could be used for diabetes diagnosis. They also stated that those with HbA1c between 42 mmol/mol and 47 mmol/mol had a high risk of progressing to type 2 diabetes and should therefore receive preventative interventions. Then, in 2010, the ADA widened the high-risk range to 39 to 47 mmol/mol, a move which has not been adopted by the EASD and IDF.
Prediabetes prevalence and risk of progression
For the individual, blood glucose and progression to type 2 diabetes result from a complex interaction between their genetic makeup and social and environmental exposures, so the prevalence of prediabetes will depend upon the characteristics of a given population and the diagnostic criteria used. However, we do know that a substantial proportion of the world’s population has prediabetes, whatever their characteristics and however it is defined. For instance, the prevalence of prediabetes among adults in China is estimated at around 50% and 38% in the USA, using ADA criteria, and around 17% in a Dutch cohort of adults aged between 45 and 75 years, according to WHO criteria.
Depending on the definition and the population, 10 to 50% of those with prediabetes will progress to overt diabetes within the next five to 10 years, but around 30 to 60% will revert to normoglycaemia within one to five years. This relatively low conversion rate may, in part, be due to the cut-off points used to define prediabetes, particularly by the ADA, as these fall within the reference limits for glucose levels in low-risk populations, particularly with increasing age. The low cut-off points might also explain why 30% of young adults in a population-based study in Liechtenstein, who had a mean BMI of around 25 kg/m2 but no other obvious risk factors, had ADA-defined prediabetic levels of HbA1c and/or IFG.
Little is known of the lifetime risk of type 2 diabetes among those with prediabetes, but in a Dutch study this was found to be highly dependent upon body size. In those with IFG and overweight/obesity (BMI > 25 kg/m2), the risk was more than 75% at age 45, but it was only around 36% for those with IFG and a BMI < 25 kg/m2.
Prediabetes has been associated with many current and future diseases, including cardiovascular disease, non-alcoholic fatty liver disease, neuropathy, chronic kidney disease, cancer and dementia, as well as all-cause mortality. These outcomes result from complex processes and people with prediabetes have different risks of developing these complications.
Interventions
Type 2 diabetes can be prevented – or at least delayed – through intensive lifestyle changes in individuals who have prediabetes. These interventions may target either individuals or population groups. However, individual intervention does not have the benefits of population-based prevention, which can shift the average risk of a population by targeting societal causes of disease, which will reduce the number of future cases.
In the Diabetes Prevention Program, the number needed to treat (NNT) to prevent one case of diabetes at the three-year follow-up was seven. The rate of conversion to diabetes among the control group was 29%. Higher NNTs have been reported in meta-analyses of trials applying lifestyle interventions in mixed settings. Most prevention trials have involved people with IGT and there is a question over whether these results are transferable to those with IFG or HbA1c-defined prediabetes. Given the lower incidence of diabetes in these two groups, it is likely that the NNT would be higher, with implications for associated costs and resources.
It’s still not known whether diabetes prevention programmes can prevent diabetes-related complications and other conditions associated with prediabetes. The authors think it likely that such programmes may well succeed in this aim, but those involved will need lifelong support and the NNT will be large if target groups are people with diabetes based on current definitions.
The way forward
The authors say that, when it comes to intervening in prediabetes, the main challenge is to strike a balance between reducing the future burden of diabetes, while also controlling costs and reducing medicalisation and stigma. There are four possible scenarios. In scenario A, the term ‘prediabetes’ is abandoned along with any interventions for those with IHG, recognising that only a minority will go on to develop diabetes. Individuals are thereby not labelled with a ‘pre-disease’ and time and resources in prevention efforts are saved. However, without intervention 25% of those with IHG will go on to develop diabetes within five years.
At the other extreme, in Scenario B, the term would continue to be used for all with IHG (around 40% of the population) and all would be offered interventions, which would reduce the incidence of type 2 diabetes by 50%, with the NNT being eight.
In scenario C, the diagnostic criteria of diabetes would be expanded to include the prediabetic range, which would result in a large increase in the numbers with the condition, many of whom would have a low risk of complications. On the other hand, many diabetes complications would also be prevented because of earlier diagnosis and treatment. It isn’t possible to calculate an NNT, as the definition of diabetes has changed, but the NNT to prevent complications would be high, given many of those diagnosed would be of low risk.
Finally, in scenario D, the risks of diabetes are estimated for all those with prediabetes and screening and intervention are offered only to those at high risk, which would reduce the incidence of diabetes by 50%, while ‘watchful waiting’ would be offered to those at low or moderate risk. In scenario D, the NNT would be four.
The financial and health impacts of these four scenarios will differ across countries and healthcare systems. Whichever is favoured, prediabetes cannot be ignored because of the present and future health burden it represents. The authors suggest the best way forward is to retain the term prediabetes, but to adopt the stratified precision medicine approach of scenario D. This recognises the varying risk among those with prediabetes and helps ensure reasonable NNTs in preventive interventions, which controls costs.
It is time to move towards stratifying those with prediabetes based upon short-term and lifetime estimated risk of diabetes, diabetes-related complications and other comorbidities. This will help inform conversations and shared decision-making with people who have prediabetes. Those at high risk will likely need lifelong support with a focus on controlling body weight. Intervening before development of overt diabetes will make reversion to normoglycaemia easier than remission of diagnosed diabetes.
As part of this approach, the authors suggest that those who have a high risk of prevalent diabetes-related complications be screened and, should complications be detected, they be treated as if they actually have diabetes, even if their glucose levels are below the diagnostic threshold. This recognises that factors beyond glucose play a role in the development of complications. Studying the viability of this screen-and-treat approach for diabetic complications among those who do not have biochemical diabetes should be a research priority.
The estimated risks discussed above rely upon the use of risk engines that can reliably identify the absolute risk of a series of outcomes, both in the future and those already established. Such risk engines, like QDiabetes, already exist but more advanced models need to be developed to fully implement the authors’ vision. The next stage will be to establish the cut-offs for estimated risk where individual-level interventions should be offered. This will require research into the cost-benefit ratio involved for individuals and society at large. Researchers and decision-makers should therefore now increase their attention and resources in this area.
The preventive interventions needed in prediabetes probably need to be long term, if not lifelong. At present, however, there is a lack of evidence for their long-term effectiveness and the trials needed to show this are costly and, inevitably, require lengthy follow-up. However, stratification by estimated risk should reduce the NNT and increase the success of the intervention. There is still the issue of adherence, so identifying interventions that people can stick to in the long term will be another important avenue of research.
Finally, policymakers should be encouraged to look at population-based approaches to diabetes prevention. These have the potential to improve the metabolic profile of the general population. These lie outside the remit of the healthcare system and require cross-sectorial collaborations to promote healthier living in society.
In conclusion, the authors emphasise that prediabetes is a risk factor, not a condition, and should be framed as such when talking to individuals, the public and decision-makers. They recommend that a stratified approach to risk and prevention be taken in prediabetes to make best use of resources to reduce the burden of diabetes in the future.
To read this paper, go to: Blond MS, Faerch K, Herder C, Ziegler D, Stehouwer CDA. The prediabetes conundrum: striking the balance between risk and resources. Diabetologia online 10 March 2023. https://doi.org/10.1007/s00125-023-05890-y
To learn more about diet and exercise interventions for diabetes, enrol on the EASD e-Learning course ‘Lifestyle intervention’: https://easd-elearning.org/courses/lifestyle-intervention/
Any opinions expressed in this article are the responsibility of the EASD e-Learning Programme Director, Dr Eleanor D Kennedy.
