A new narrative review led by researchers at UCL and Birmingham Children's Hospital has found there is a lack of evidence to demonstrate the effective use of continuous glucose monitors (CGMs) in people not living with diabetes (PNLD).
In the study, published in Diabetic Medicine, researchers conclude there is currently little published evidence on how accurate CGMs are in measuring blood glucose levels in PNLD, nor sufficient evidence of what the health benefits or utility such information would provide. The researchers also suggest CGMs may have unintended adverse health effects in this user group and are now calling for better regulation.
In recent years, CGMs have helped transform the treatment of type-1 diabetes and have been used with good effect by people with type-2 diabetes who are on insulin therapy. The devices, originally developed for those with diabetes, are attached to the body and monitor blood glucose levels in real-time, transmitting the information to an insulin pump in people with type-1 diabetes that delivers the correct amount of insulin required to keep blood glucose levels within a defined range2.
For those living with diabetes, CGMs are less burdensome and offer real-time trends in glucose changes compared to older testing methods, such as regular finger prick blood tests.
However, more recently CGMs have been increasingly marketed to PNLD for uses without a medical indication, such as health, wellness and lifestyle management, with some companies providing advice around how to manage glucose levels or 'glucose spikes'.
Explaining the current interest in CGMs, senior author and dietitian, Dr Adrian Brown (UCL Division of Medicine), said: "Several health companies are now using CGMs as part of paid-for programmes designed to give people personalised nutrition information.
Some claim to tailor a person's diet and activity to help keep blood glucose within 'normal' levels. But what constitutes a normal blood sugar level varies between individuals and in the same individuals at different times, and CGM accuracy varies between CGM models."
Dr. Adrian Brown, Senior Author and Dietitian, Division of Medicine, University College London
"With this in mind, we wanted to see what research had been undertaken into the use of CGMs in people who do not have diabetes."
In this narrative review, the team from UCL and Birmingham Children's Hospital searched online databases (PubMed, Medline, Embase and the Cochrane Library), in the years 1980-2023, for studies examining aspects of CGM utility and performance in PNLD. They found 25 relevant studies.
Researchers then looked to establish what, if any, evidence there was on how effective CGMs were at measuring glucose levels, glucose variability and elevated glucose levels in PNLD, as well as any research on the influence of CGM use on eating behaviours.
The review concludes that there is a lack of consistent and high-quality evidence to support CGM use in PNLD. For instance, there is little evidence to show how accurate CGMs are at measuring blood glucose levels or detecting changes in PNLD, nor sufficient research into the value and utility of the CGM data obtained in PNLD.
The review also found evidence that CGM use in PNLD could cause anxiety about what is normal in terms of diet and blood sugar levels. Researchers say this could present a risk of developing eating disorders, such as orthorexia (an unhealthy obsession with eating 'pure' food).
An earlier separate review by the authors, referenced in this paper, looked at the regulation of CGM in people with diabetes, and found there was no specific statutory guidance available.
Dr Brown, who is based at the UCL Centre of Obesity Research, said: "While there are some benefits to personalised health programmes offered by commercial providers, including getting people to think about what and how much they're eating, the fact is that we don't have the same health outcome data for CGM use in people not living with diabetes.
"This review unpicks the idea of 'abnormal glucose', the accuracy of CGM data, and the effect on behavioural change when CGMs are used by people without diabetes.
"At the moment there are big questions about how much guidance these health companies are giving customers to help them interpret their glucose data, and what scientific evidence this guidance is based on. This can leave the customer to interpret what the variations in their blood glucose mean, posing a risk that they misinterpret the data and avoid certain foods unnecessarily."
Co-author John Pemberton, a Specialist Paediatric Diabetes Dietitian from Birmingham Children's Hospital who sits on the International Federation of Clinical Chemistry (IFCC) and Laboratory Medicine Working Group on CGM, said: "CGMs with an accuracy within 20% of actual blood glucose levels, at least 95% of the time, represent current market leading performance, and are extremely helpful for people with diabetes to make daily treatment decisions.
"However, the regulations for CGMs for people living with diabetes are ambiguous both nationally and internationally, making it difficult to know if available CGMs meet this level of accuracy. This is why the IFCC CGM group are pushing for an international standard.
"For individuals not living with diabetes, the situation is even more uncertain. We have little robust information on whether CGMs achieve the required accuracy in this population.
"Normal glucose levels for individuals without diabetes range from 3.3 to 7.8 mmol/L and the most accurate CGMs have a 20% accuracy margin most of the time. This means that the most accurate CGM devices will display readings between 2.6 and 9.4 mmol/L, even when the actual glucose level is within the normal range. Such discrepancies can lead to unintended stress and potential psychological and behavioural implications.
"While CGMs for people not living with diabetes show promise, the accuracy, regulatory standards, and psychological impacts of false highs and lows are not well understood. Despite this, CGMs are being heavily promoted without mention of these issues."
Could vitamin D be the missing link in managing type 2 diabetes? Recent study weighs in