top of page

Essential tips for using Continuous Glucose Monitoring (CGM)

In a series of previous blogs on mysportscience, the role of blood glucose was discussed and we highlighted what insights Continuous Glucose Monitoring (CGM) can provide athletes now and possibly in the future. In this blog, Dr Nicola Guess highlights the multifactorial nature of glucose metabolism, challenging the notion of straightforward causal links between food intake and glycemic responses.

Pattern of glucose spikes

In previous blogs Dr Mike Riddell and I discussed opportunities and also some limitations of CGM.

There are many ways in which we can look at CGM data. We could simply look at a value, or an average over 24 hours. We could look at the glucose variability (GV), or simply count the numbers of peaks and troughs, or perhaps the time spent in a certain zone, or time spent above or below a predetermined value. We could look at the slope of a peak, we could average glucose values during sleep or during exercise, or just in response to one meal. The possibilities are endless.

Like many wearables, CGM produces a lot of data and the interpretation of that data is key to its potential usefulness or even harmfulness. This will be the topic of this guest blog by Dr Nicola Guess, a Registered Dietitian with a PhD in the dietary management of prediabetes from Imperial College London who currently works at the University of Oxford. In this blog she will talk about common misconceptions and misinterpretations. She will also discuss the modifying effects of exercise on glucose responses and some of the essential differences in glucose behaviour between those individuals living with diabetes and healthy individuals.

We don't know what "normal" is yet on continuous glucose monitors!

A common first question from someone using CGM is: “my value is 7.8 mmol/L [140 mg/dL], is this good or bad?” The use of CGM devices in individuals (including athletes) without diabetes is very recent, and while there's emerging data, we can’t define any specific cut-offs for “good” and “bad”. CGMs have been used for a lot longer in type 1 diabetes, and it’s only in the past couple of years that we have been able to establish some of the glycaemic targets people with type 1 diabetes should aim for on a CGM. This has been possible because we now have large datasets of people with type 1 diabetes, so we can determine the glycaemic parameters from a CGM and see how they relate to the risk of things like hypoglycaemic episodes and diabetic complications. We have zero useful data on CGM-derived glycaemic measures in populations without diabetes and health outcomes! So how can we know for sure what glucose is “ok” on a CGM in a person without diabetes? We don’t, and we won’t for a while.

We have zero useful data on CGM-derived glycaemic measures in populations without diabetes and health outcomes!

The best we can do is look at the CGM profiles of people with a “healthy” HbA1c.

What is HbA1c?

The term HbA1c refers to glycated heamoglobin. It develops when haemoglobin, a protein within red blood cells that carries oxygen throughout the body, joins with glucose in the blood, becoming ‘glycated’. By measuring glycated haemoglobin (HbA1c), clinicians are able to get an overall picture of what our average blood sugar levels have been over a period of weeks/months.





Below 42 mmol/mol

Below 6.0%


42 to 47 mmol/mol

6.0 to 6.4%


48 mmol/mol and over

6.5% and over

*the cut-off point for pre-diabetes can vary between countries. For example, pre-diabetes in the USA starts at 5.7%.

Observations from studies investigating the CGM profiles of people with a “healthy” HbA1c (under 5.7%), CGM-derived glucose readings typically show infrequent excursions above 7.8 mmol/L (140 mg/dL), albeit with some variability across age groups. However, caveats exist, including potential alterations in behavior during study participation, which might lead to underestimation of typical glucose peaks. Despite these challenges, data from multiple studies hint that occasional post-meal excursions into the 9-11 mmol/L (162-198 mg/dL) range may be common among individuals with satisfactory HbA1c levels, possibly reflecting a normal aspect of human glucose metabolism.

We also don’t know if glycaemic variability (peaks and troughs) in people without diabetes even matter, and if so, how much. Instinctively it might seem that of course glucose going up and down a lot must be a bad thing…? But we don’t actually have evidence of this yet. Studies conducted in vivo (in humans) and in vitro (in cells) primarily focus on the type of glucose variability you get in diabetes, where they are much, much higher (and usually sustained for longer) and the drops are lower. Such studies demonstrate the potential for high glucose variability to induce endothelial damage and other complications – but do not accurately reflect glucose variability in individuals without diabetes, where glucose fluctuations are typically less pronounced and of shorter duration.

Despite claims regarding the benefits of stabilising blood glucose levels, empirical evidence linking glucose variability to health outcomes, particularly in normoglycemic individuals, remains scant. While observational studies suggest weak associations between glucose fluctuations and factors like hunger and energy intake, randomised controlled trials fail to establish a significant relationship between glycaemic peaks and dips and appetite regulation. The complex interplay of biological, psychological, and environmental factors influencing appetite suggests that glycaemic variability alone may not play a meaningful role in driving appetite responses.

Glycaemic variability alone may not play a meaningful role in driving appetite responses.

A transient high glucose doesn't mean you have prediabetes or type 2 diabetes

Impaired glucose tolerance (prediabetes) is diagnosed by measuring a person’s blood glucose 2 hours after a drink with 75g glucose in it. If a person’s glucose is 7.8mmol/L (140 mg/dL) or above 2 hours after the drink, we call this impaired glucose tolerance. However, this does not mean that if your glucose gets to 7.8mmol/L (140 mg/dL) at any time, that you have prediabetes. It is completely normal for your blood glucose to get to 7.8mmol/L (140 mg/dL) 30 mins after a drink of 75g glucose or a similar carbohydrate load. The point is that it comes down afterwards.

Glucose peaks versus chronically high glucose

The same is true with diabetes. We diagnose diabetes when fasting glucose is 7.0 mmol/L (126 mg/dL) or more, or when 2-hour glucose is 11.0 mmol/L (198 mg/dL) or more. But this doesn’t mean that if your glucose ever gets to 7.0mmol/L (126 mg/dL) or 11.0 mmol/L (198 mg/dL) that you have diabetes!


In fact, calling a glucose spike up to e.g. 7.8mmol/L (140 mg/dL) “prediabetes” or a spike up to 11.0 mmol/L (198 mg/dL) “diabetes” completely misunderstands the nature of these conditions. Both prediabetes and diabetes are characterized by chronically elevated glucose.


If a person’s glucose is still at 7.8mmol/L (140 mg/dL) 120 minutes after the glucose drink it’s an indication their glucose has been at least 7.8mmol/L (140 mg/dL) for the better part of 90 mins. Likewise, if a person’s fasting glucose is 7.0mmol/L (126 mg/dL) it’s probably been this high most of the night. And if their 2-hour glucose is 11.0 mmol/L (198 mg/dL), their glucose is probably elevated