The measure of it – Time in range

Dr Angela Murphy discusses diabetes control in the 21^st century, specifically time in range.

After the discovery of insulin in 1922, people with diabetes (PWD) were living longer and, thus, diabetic complications started to emerge. The goal in managing diabetes started to focus on preventing these complications and improving overall quality of life. With the advent of large-scale clinical trials, the relationship between glucose levels and risk of complications was revealed.

The Diabetes Control and Complications Trial (DCCT), which involved Type 1 diabetes patients, released its results in 1990. It conclusively showed that improved glucose control reduced the risk of microvascular (small blood vessel) complications of the eyes, kidneys, and nerves.

Large trials done in Type 2 diabetes patients, such as the United Kingdom Prospective Diabetes Study (UKPDS), confirmed these reductions in microvascular complications with good diabetes control.

Initially, the data didn’t show the same benefit in macrovascular (large blood vessel) complications, viz: coronary artery disease, stroke, and peripheral vascular disease where blood supply to the feet is blocked. However, trials that followed-up patients beyond 15 years did show that there was indeed benefit. We now had proof that good glucose control prevented complications. So, what constitutes good diabetes control?  How do we measure that?

Blood glucose measurement

The impetus to test glucose in the early days of diabetes was centred on diagnosis. Hence, initial urine test kits would just have established if glucose was present or not.

Then PWD wanted to assess the degree of improvement in their glucose levels using the different treatment regimens available. Another strong reason to test would have been to establish hypoglycaemia.

Blood glucose testing was greatly improved in the mid 1960s with the invention of the Ames dextrostix and gradually home glucose testing kits became available.

When continuous glucose monitoring (CGM) devices were released, even more accuracy was expected. Blood glucose targets were partly determined from knowing what the non-diabetic blood glucose range is, but also from the data of the trials. There are three parameters that are used for these targets: fasting blood glucose, post-prandial (two hours after a meal) blood glucose, and HbA1c.

HBA1C – glycated haemoglobin

In 1968, Iranian born Samuel Rahbar started studying the haemoglobin (Hb) molecule (which carries iron and oxygen in our blood). By chance he came across an unusual variant in diabetic blood and went on to work at the Albert Einstein College of Medicine with Helen Ranney. They found that their diabetic Hb matched the previously described subtype of HbA and so was named HbA1c.

HbA1c is formed when excess glucose in the blood attaches to the haemoglobin molecule, a process called glycation. Red blood cells are renewed on average every three months, so HbA1c is regarded as an average of blood glucose control over a three-month period. In fact, it was HbA1c that was used in the DCCT and UKPDS trials to show that good glucose control prevented diabetes complications. Without HbA1c, this would have been nearly impossible to demonstrate.

Home glucose monitoring

In this era of data overload, it’s truly useful to be reminded why we advocate blood glucose testing in the management of diabetes. It should not just be about going through a routine to provide the doctor with readings at the consultation, but also to empower the PWD.

In 2011, the Structured Testing Program Study showed that getting PWD to do structured testing in the three days before their doctor’s visit improved their HbA1c by 0,3%. There was no medication change, just behavioural change.

The participants in the study used the AccuCheck 360’ tool (below) and did a seven-point profile for those three days. This involved testing before each meal, two hours after each meal and at bedtime. I have adapted this tool for my clinical practice and use it frequently. It helps me advise patients with medication changes, but it’s just as valuable for the patient to see what else influences the glucose levels. This can range from food to medication, activity, illness and even stress. It is particularly useful when patients are new to the practice and their diabetes regimen needs to be assessed.

Continuous glucose monitoring

The first generations of CGMs approved by the Food and Drug Administration (FDA), beginning in 1999, were able to provide significant clinical benefits as an adjunct to standard self-monitoring of blood glucose.  These are the machines of many a PWD’s dreams: a way of seeing the blood glucose at any time of day or night without having to open a conventional glucometer and prick a finger.

As the CGM devices became more advanced, they not only showed the current glucose reading, and of course the tracing of where the glucose had been but could predict where the glucose would go. In this way, PWD could be forewarned of a hypoglycaemia or hyperglycaemia and take appropriate action to avoid these. When this type of CGM technology works in tandem with insulin pumps, we see the makings of an artificial pancreas.

Time in range

What data from CGM showed acutely is that we cannot always rely on average blood glucose levels, even HbA1c, to fully assess overall diabetic control. Averages do not show the extent of the high and low glucose readings.

Let me explain. If there are three blood glucose values of 6,0mmol/L, then obviously the average blood glucose is 6,0mmol/L. However, three readings of 12mmol/L, 2mmol/L and 5mmol/L will also give an average of 6,0mmol/L and yet only one reading is in the target range.

Figure 3 shows how all HbA1c’s are truly not equal. The same HbA1c of 7,0% can have completely different glucose profiles. This variation in glucose levels is called glucose variability. CGM demonstrates patterns of glucose over a 24-hour period in detail so the swings in blood glucose levels are easily seen. There is strong evidence to show that increased glucose variability predicts the risk of hypoglycaemia. Specifically, it predicts severe hypoglycaemia in Type 1 diabetes and non-severe in Type 2 diabetes.

Severe hypoglycaemia is defined as low blood glucose <4mmol/L and the PWD requires assistance to treat the low glucose. The more frequently the blood glucose levels swing from highs to lows, the higher the glucose variability. There is concern that this variability can damage blood vessels and thus, may be implicated in diabetic complications.

Glucose targets

Based on data from all the large diabetes trials over the years, we can set targets for good diabetes control. This is not a one-size fits all range. Age, duration of diabetes, presence of complications, risk of hypoglycaemia and pregnancy all affect the target blood glucose levels.

In older PWD who have diabetic complications, particularly of the heart and kidneys, glucose levels are slightly higher than a young, newly diagnosed PWD. Table 1 shows advised targets for FBG, PPG, HbA1c.

TABLE 1 – BLOOD GLUCOSE TARGETS FROM SOCIETY OF ENDOCRINOLOGY AND METABOLISM SOUTH AFRICA 2017

In 2019, the International Consensus in Time in Range (TIR) was released and defined the concept of the time spent in the target range between 4 and 10mmol/L while reducing time in hypoglycaemia, for patients using CGM.

Several studies have now shown a good correlation of TIR with HbA1c. In one, more than 90% participants with a TIR of >80% has HbA1c values of ≤ 7.0%.

At present, TIR is only verified with the use of CGM. Intermittent testing, even doing seven-point profiling does not seem to be as predictable. Several medical aids will now consider reimbursement for CGM devices for Type 1 diabetes patients, which will significantly increase the use of CGM in South Africa.

Summary

To achieve good diabetes control, we try to get as close to physiological glucose levels as is safe which has been proven to decrease both microvascular and macrovascular complications. Good control is not only a good average but also stability of glucose levels over time. It is this latter attribute which is measured with Time in Range and which may become the most important of all glucose measurements in the future.