Glucose monitors stuck to the arm may be inaccurate when it comes to telling people if they are at high risk of developing diabetes. The devices can suggest different results for whether someone has healthy blood sugar readings from one day to the next, a study has found.
The inconsistencies suggest that more research is needed in people without diabetes before the monitors can be used to give them any health information, says at the Weizmann Institute of Science in Rehovot, Israel.
Continuous glucose monitors are disc-shaped devices, usually placed on the upper arm, which send blood sugar readouts to people’s smartphones.
Advertisement
They were developed so that people with diabetes didn’t have to carry out frequent “finger prick” blood tests to check their blood sugar hadn’t gone too low or high. But some firms are now selling these monitors, along with dietary advice services or supplements, as a way for people without diabetes to lose weight or otherwise improve their health, although there is little good evidence supporting this approach.
One way that the monitors could, in theory, be helpful is in identifying people with type 2 diabetes or “prediabetes”, a state where blood sugar levels are somewhat higher than average, but not high enough to indicate type 2 diabetes.
About 10 to 20 per cent of people with prediabetes progress to type 2 diabetes over the following decade, but people can generally reduce the risk of this happening by losing weight.
Shilo’s team investigated continuous glucose monitors as part of a large ongoing health study of people aged between 40 and 70, who were asked to wear a monitor for two weeks.
The researchers obtained a “fasting glucose” measure – taken in the morning when people had gone at least 8 hours without food – for 8300 people, over an average of one week. They classed the readings as either normal, in the prediabetes range or above the threshold that indicates diabetes.
The team found that there was great variability in people’s results from day to day. While most people’s fasting glucose readings on the first day of the study were normal, 40 per cent of these participants had a further reading that was in the prediabetes range, and 12 per cent of those whose first reading indicated prediabetes then had completely normal values for the rest of the study period.
Shilo says further studies are needed to establish whether continuous glucose monitors can classify someone as having prediabetes via a more complex measure, for instance if their glucose level is above a cut-off for a certain percentage of the time, says Shilo.
at the University of Oxford says some people without diabetes who are using the monitors may wrongly think a single high value shows they have prediabetes. Doctors are more likely to consider a range of measures, including a different kind of blood test that reflects average blood sugar values for the past few months, called haemoglobin A1c, she says.
“We need ways to identify people who are on their way to type 2 diabetes early,” says Guess. “The idea that it would be feasible or valuable for people to wear a CGM [continuous glucose monitor] for two weeks as a screening tool for prediabetes seems a little bit absurd.”
Nature Medicine
