What is hypoglycaemia?

The wider use of continuous glucose monitoring (CGM) raises questions about the significance of low glucose readings. According to findings presented at the recent Advanced Technologies & Treatments for Diabetes (ATTD) conference, there is a difference between sensor-detected and person-detected hypoglycaemia, but the clinical and personal significance of this difference needs to be explored further. Dr Susan Aldridge reports.

Although hypoglycaemia remains a major contributor to the burden of diabetes, there is still much to be learned about the impact of low glucose on the health and wellbeing of people living with the condition. Even the definition of hypoglycaemia is still a matter for debate. Simon Heller, Professor of Clinical Diabetes at the University of Sheffield, noted that the International Hypoglycaemia Study Group has argued the case for three levels to reflect the complexity of hypoglycaemia. These are Level 1 (‘alert’) corresponding to a glucose level of 3.9 mmol/l or below, as set by the American Diabetes Association in 2005, which has long been widely accepted as the definition of hypoglycaemia; Level 2 (‘serious’) occurs at a glucose level of 3.0 mmol/l, while Level 3 (‘severe’) is a hypo involving cognitive impairment to the extent of requiring the assistance of a third party. 

This has led to ongoing debate between stakeholders, including the regulatory authorities, over what constitutes a ‘minimally important, clinically meaningful difference’ in hypoglycaemia, which is important when looking at the impact of interventions.  

Whatever the outcome of these discussions, hypoglycaemia is important to those who experience it. In a study funded by the Juvenile Diabetes Research Foundation (JDRF), people with diabetes were asked how much they were willing to pay for various improvements in their condition. For a 0.5% improvement in HbA1c, which most would agree is the minimum clinically important difference, they would pay $5 a month. However, for a reduction from one serious hypo and five to seven mild to moderate hypos per week to just two to four mild to moderate lows per week, they’d be willing to pay as much as $73 a month. 

So research into hypoglycaemia should be a priority. “The importance of more evidence is clear and will be provided by the work of Hypo-RESOLVE,” said Professor Heller. Hypo-RESOLVE, a consortium of industry, academic and other partners, including people with diabetes, was set up in 2018 to investigate hypoglycaemia and its impact. In the words of members of the Hypo-RESOLVE Patient Advisory Committee, ‘any fall in day-to-day mild hypos would make the most difference to my life’ and ‘halve my non-severe lows to no more than one a week’. 

Sensor versus personal hypo detection

One of Hypo-RESOLVE’s work packages involves looking at the significance of low glucose detected by CGM. People with diabetes don’t always know when they are having a hypo, but their CGM does. Pratik Choudhary, Professor of Diabetes at Leicester Diabetes Centre, showed a trace where the subject was a concerning 23% time below range. “Our response as clinicians is to look at what’s causing this,” he said. “Are they taking too much basal insulin; are they over-correcting at night? To us, it’s numbers, but to the person, it could just mean annoyance at the alert noise caused by the technology.”

CGM in people without diabetes shows that it is quite usual to get some readings in time below range, so are these really harmful? For a person with diabetes, these readings could correlate with impaired awareness of hypoglycaemia (IAH), if they are not experiencing any symptoms. However, at the moment, there isn’t the data to make this connection. “We also know that there is a link between hypos and inflammation and vascular risk, but again we don’t yet know the correlation between time below range and these risks,” said Professor Choudhary. 

The difference between sensor and person-reported hypos was demonstrated by a study from Denmark, which used a blinded CGM with 100 participants. They were asked to report any hypos they experienced and these were compared with those detected by the sensor. While there were five to seven low glucose sensor events per week, there were only one to two self-reported hypos. 

This finding – of 70% of sensor hypos being asymptomatic – was repeated in the InRange randomised controlled trial, which was actually designed to compare insulin glargine and insulin degludec. This was the case at both 3.9 mmol/l (IHSG Level 1) and at the lower glucose level of 3.0 mmol/l (IHSG Level 2), so it’s not that people are better at detecting the lower glucose levels. “We’re seeing a story here that a lot of low events are not symptomatic,” said Professor Choudhary. Therefore, the Hypo-RESOLVE group is now trying to look at this in a more organised way with the Hypo-METRICS trial.

The Hypo-METRICS trial

Hypo-METRICS is a multicentre, multinational observational study exploring CGM hypos to further understanding of the clinical, psychological and health economic effects of hypoglycaemia. The trial involved 602 participants with type 1 or type 2 diabetes, all on insulin and having reported at least one hypo in the last three months. They wore a blinded Freestyle Libre 2 sensor and continued with their usual glucose monitoring, reporting any symptomatic hypos on the dedicated Hypo-METRICS app.

This meant the researchers could map whether or not sensor and participant-reported hypos matched up. The study yielded 40,962 study days, equivalent to 102 patient years, or just under one million hours, of CGM data. Overall, there were around 15,000 symptomatic hypos and 30,269 sensor-detected Level 1 hypos, 7,424 Level 2 hypos and just 302 at Level 3. 

The researchers looked at whether those with IAH were less likely to experience a sensor-detected hypo than those with normal awareness and found no difference. And when it came to the comparison between sensor and symptomatic events, the overlap was 8,259 events, which were experienced as symptomatic and reported on the app and also detected by CGM. “This means that for around half of the 15,000 person-reported events, glucose was actually above 3.9 mmol/l,” Professor Choudhary explained.    

Do asymptomatic hypos matter?

There is evidence that low glucose, whether or not it is symptomatic, is linked to adverse outcomes, such as seizures and cardiac arrhythmia. And there are new findings from Hypo-RESOLVE that show an association with inflammation.  

So while many episodes of sensor-detected hypoglycaemia are asymptomatic and have little effect on patient experience, they may be causing silent damage and have an impact on long-term outcomes. Recognised hypos, where an alarm goes off, whether or not there are symptoms, do have an adverse effect on patient reported outcomes, irrespective of the glucose level, according reports on the Hypo-METRICS app. 

“A CGM hypo is not the same as a patient-reported hypo,” said Professor Choudhary. “So my plea to academic and industry partners is to ensure that both be reported. We now may be able to use these data to educate people with diabetes further about hypoglycaemia and its effects.”  

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