Continuous Glucose Monitoring
This is an invaluable technique that is becoming more widely available as technology improves (wireless, and sensors that can reliably measure glucose levels for 4–5 days, in some cases longer) and costs fall. The subcutaneous sensor measures interstitial, rather than blood, glucose, and there is delay of about 15 minutes, of which half is physiological and half due to instrumental processing delays. This is not critical when CGM is used for diagnostic purposes, but is very important when used in real-time monitoring (Fig. 4.4) or when CGM data are transmitted to insulin pumps (sensor-augmented pumps; see Fig. 4.3). Nevertheless, patterns of glycemia are easier to detect than with even frequent home blood glucose monitoring; where there are rapid changes in glucose levels, small differences in timing of insulin injections and fingerprick testing can give an impression of greater variability than there actually is. Long-term use of CGM can lower HbA1c by about 0.5% in adults 25 years or older (though not in children and adolescents), with a simultaneous decrease in severe hypoglycemia [7].

Associations between CGM and other measures of glycemia are emerging, and if CGM is now considered the gold-standard test, the reverse correlations with self-monitoring of blood glucose (SMBG) are important for everyday clinical practice.
- CGM and SMBG results are strongly correlated – SMBG is still valuable.
- SMBG 90 min after a meal correlates well with postprandial glycemic exposure on CGM; we should probably recommend 90-min rather than 2-hour postprandial monitoring.
- Fasting glucose levels are only weakly associated with other glycemic measures.
- Preprandial glucose measurements impact more than postprandial measurements on HbA1c, especially in type 2 diabetes [8].
A combination of home blood glucose monitoring and CGM frequently identifies characteristic patterns (Fig. 4.5).

Nocturnal Hypoglycemia
This is very common. In a large-scale study of CGM, nocturnal hypoglycemia (glucose < 3.3 mmol/L, 60 mg/dL) was equally frequent with CSII and MDI, and increased with decreasing HbA1c. Episodes lasting 20 min or more occur about every 14 days, and nearly one-quarter of all episodes last 2 hours or more. This is another caution against the automatic assumption that ‘modern’ insulin technology automatically reduces the most serious complication of type 1 diabetes [9]. Encourage patients to reduce their bedtime long-acting insulin, especially where there has been previous physical activity or alcohol intake. Small changes can have dramatic effects.
Fasting Hyperglycemia
Abolishing nocturnal hypoglycemia may still not help fasting hyperglycemia, or high variability of fasting glucose levels. Poorly controlled type 1 diabetes (and puberty) is associated with the dawn phenomenon, exaggerated surges of growth hormone in the early hours of the morning, leading to delayed insulin resistance and fasting hyperglycemia. This is easily corrected in pump-treated patients. Declining insulin levels from an inadequate dose of long-acting bedtime insulin, which will give a similar glucose pattern, is uncommon but easily correctable.
Where there is no evidence for nocturnal hypoglycemia and the dawn phenomenon, encourage slow up-titration of bedtime long-acting insulin. Modest dose increases (about 10–15%) often significantly improve control.
Postprandial Hyperglycemia
This is difficult to manage with MDI. The action of even rapid-acting analogues is too slow to counteract peak postprandial glucose levels, and increasing the dose may simply increase the tendency to early postprandial hypoglycemia. The slower onset but longer action of soluble/regular insulin may help some people at particular times of day, and taking prandial insulin 15–30 mins before a high-carbohydrate meal may help, even though this may be considered old-fashioned practice. A more rigorous approach to carbohydrate counting and flexible insulin dosing can be of real value, and modern insulin pumps have several ingenious strategies to help manage postprandial peaks, for example dual-wave boluses. For really motivated patients taking high glycemic index carbohydrate meals, about one-third of the mealtime bolus can be taken before the meal, the remainder after.
New Developments
New-generation long-acting insulin analogues are in clinical trials. Insulin degludec (NN1250, NovoNordisk) is a soluble insulin basal analogue suitable for three-times weekly dosing, and gives glycemic control and hypoglycemia rates similar to once-daily glargine. A combination of degludec with aspart (NN5401, a soluble insulin analogue combination, SIAC) may have a basal-bolus profile. SAR161271 (Sanofi-Aventis) will be given daily.
For more information and to purchase this book, just follow this link:
https://www.wiley.com/WileyCDA/WileyTitle/productCd-1444333852.html
David Levy, MD, FRCP, Consultant Physician, Gillian Hanson Centre, Whipps Cross University Hospital; Honorary Senior Lecturer
Queen Mary University of London London, UK
This edition first published 2011, © 2011 by David Levy. 1st edition 1998 (Greenwich Medical Media/Cambridge University Press) 2nd edition 2006 (Altman Publications)
Diabetes In Control. A free weekly diabetes newsletter for Medical Professionals. News and information for Medical Professionals.