A time honoured approach to reducing the frequency of nocturnal hypoglycaemia has been to counteract the effect of insulin by ensuring that patients eat a bedtime snack. This seemed particularly important for children who go to bed early and sleep for many hours between their evening insulin dose and breakfast on the following day. It clearly makes sense for all individuals taking insulin to measure their blood glucose before bed and to take additional food if their blood glucose is low. However, the extent to which protective eating can prevent nocturnal hypoglycaemia in patients taking intermittent injections of insulin is limited.
Box 4.3 Potential remedies for problematical nocturnal hypoglycaemia
1. Long and rapid-acting insulin analogues.
2. ^-agonists (terbutaline or salbutamol).
3. Appropriate snacks:
- uncooked cornstarch
- high protein foods.
The effectiveness of different approaches has generally been assessed by measuring the effectiveness of different snacks in preventing biochemical or symptomatic episodes of hypoglycaemia. Before the advent of continuous blood glucose monitoring, such studies demanded regular blood sampling for measurement of glucose and since this generally required admission to hospital for study, these investigations were not exactly examining a typical clinical situation. Furthermore, the number of subjects studied has generally been relatively small, and although sufficient to measure changes in blood glucose concentration or frequency of biochemical hypoglycaemia, the studies have been inadequately powered to determine effects on the frequency of severe episodes. Some studies have examined the potential of carbohydrate foods that are absorbed slowly and thus able to counter the hypoglycaemic effect of insulin over longer periods. Uncooked cornstarch, which has a very low glycaemic index, has been studied intensively, particularly because it is used successfully to prevent severe hypoglycaemia in glycogen storage diseases (Goldberg and Slonim, 1993). In one study, when young people were given uncooked cornstarch incorporated into a normal bedtime snack, the incidence of symptomatic and biochemical nocturnal hypoglycaemia was three-fold lower (Kaufman and Devgan, 1996). A blinded randomised trial reported a similarly lower frequency of nocturnal episodes (Kaufman et al., 1997). Other work has examined the effect of snacks rich in protein or fat.
Different types of snack were compared against placebo in a trial using a crossover design in 15 adults with type 1 diabetes (Kalergis et al., 2003). When patients retired to bed with blood glucose greater than 10.0mmol/l, nocturnal episodes were not observed. At a pre-bedtime glucose below 7.0 mmol/l, a high protein snack prevented any episode of nocturnal hypoglycaemia by contrast with either a standard snack or one containing cornstarch (Figure 4.4). Thus the clinical data measuring the effectiveness of cornstarch are
Pl S CS Prot 7-10 mmol/L Bedtime blood glucose category
Figure 4.4 Effect of snacks and bedtime blood glucose concentration on frequency of nocturnal hypoglycaemia. Open box = neither nocturnal hypoglycaemia nor morning hypoglycaemia; solid black box = hypoglycaemia; striped box = morning hypoglycaemia only. Snack type: Pl = placebo; S = standard diet; CS = cornstarch; prot = protein conflicting. Since uncooked cornstarch is not easy to prepare in a digestible form, it is not surprising that it is not used widely to prevent nocturnal hypoglycaemia.
An alternative approach to dietary supplements is to reduce the rate of absorption of carbohydrates using a disaccharidase inhibitor such as acarbose. Three studies of these agents have examined their effects in patients with type 1 diabetes. McCulloch et al. (1983) studied the effect of acarbose on the risk of overnight hypoglycaemia, and found that the risk of symptomatic nocturnal hypoglycaemia was lowered by 39%. Taira et al. (2000) reported similar benefits using voglibose. However, a recent study reported no benefit of acarbose over placebo when both pharmaceutical and snacking interventions were compared with respect to their effect on preventing hypoglycaemia (Raju et al., 2006). In the light of these limited and conflicting data it seems unlikely that acarbose will never be widely used, particularly as sucrose-containing products cannot be used as a treatment for hypoglycaemia when acarbose is being taken.
There are indications that jS-agonists may have some use in reducing the risk of nocturnal hypoglycaemia. For some years inhaled terbutaline has been proposed as a method of elevating blood glucose. In the early 1990s, Wiethop and Cryer (1993) demonstrated that its oral or subcutaneous delivery following induced hypoglycaemia, led to a rise in blood glucose compared to placebo, an effect that lasted for some hours. More recently, Wright and Wales (2003) reported that children with type 1 diabetes who were receiving treatment for asthma had fewer episodes of nocturnal hypoglycaemia when compared to a non-asthmatic group of children in a survey lasting three months, and they implicated a beneficial effect of jS-agonist therapy. Raju et al. (2006) also compared the effect of inhaled terbutaline with other therapeutic remedies such as cornstarch, standard snacks and acarbose on nocturnal blood glucose levels in patients with type 1 diabetes. They found that terbutaline prevented nocturnal hypoglycaemia in all 15 subjects. However, although this treatment offered the greatest protection against hypoglycaemia at night, it also led to the highest fasting blood glucose among the different remedies, indicating that additional work needs to be done to establish this treatment as a realistic therapeutic option.
Timing and Type of Insulin, Including Insulin Analogues
The introduction of insulin analogues with pharmacokinetic properties that bear more resemblance to physiological insulin profiles in non-diabetic individuals highlighted the potential of such preparations to lower the risk of hypoglycaemia. Over a full 24 hours, the overall frequency of symptomatic hypoglycaemia in trials of rapid-acting insulin analogues has been modestly lower than with conventional insulins, but a consistent finding has been a lower rate of nocturnal hypoglycaemia. It appears that the tendency of conventional soluble insulin to self-associate into hexamers at therapeutic concentrations leads to increasing plasma insulin levels with repeated injection. Since rapid-acting insulin analogues separate into single molecules much more readily, accumulation of insulin is less likely and the risk of nocturnal hypoglycaemia is subsequently lower. The frequency of nocturnal hypoglycaemia observed in clinical trials has been variable. Rates of nocturnal hypoglycaemia have been over 50% lower in some trials involving patients with type 1 diabetes with strict glycaemic control (Heller et al., 1999; Heller et al., 2004), and although this has generally been demonstrated for symptomatic episodes, these data might reflect a lower frequency of severe hypoglycaemia (Holleman et al., 1997).
The long-acting analogues - insulins glargine and detemir - which provide basal insulin replacement, also appear to lower the risk of nocturnal hypoglycaemia. They have a longer duration of action when compared to isophane (NPH) insulin, together with less of a peak in their time-action profile and a more consistent duration of action (Barnett, 2003). Both of these properties probably contribute to the lower rates of nocturnal hypoglycaemia that have been observed during clinical trials. Relative risk reductions of around 30% have been reported for long-acting preparations in trials involving patients with type 1 (De Leeuw et al., 2005; Pieber et al., 2000) and with type 2 diabetes (Hermansen et al., 2006; Yki-Jarvinen et al., 2000).
The combination of both rapid and long-acting insulin analogues might be expected to have a particularly powerful effect in lowering the risk of nocturnal hypoglycaemia. In the few studies comparing combinations of insulin analogues to conventional insulins this appears to be the case. Hermansen et al. (2004) reported a 55% lower rate of symptomatic nocturnal hypoglycaemia when using an insulin detemir/aspart combination as basal-bolus therapy in patients with type 1 diabetes, which was accompanied by a modest but significant fall in HbA1c of 0.2%. Ashwell et al. (2006) observed a fall in HbA1c of 0.5% using insulin glargine and lispro in a basal-bolus regimen in patients with type 1 diabetes, while nocturnal hypogly-caemia was 44% lower in frequency. However, nocturnal hypoglycaemia was not eradicated, leading to the conclusion that although insulin analogues may help to reduce the side-effects of insulin therapy, they do not approach the requirements of therapeutic insulin delivery.
Continuous Subcutaneous Insulin Infusion (CSII)
Since nocturnal hypoglycaemia is largely the result of inadequate basal insulin replacement, one would expect that the most effective method of basal insulin delivery currently available, CSII, could limit the frequency of nocturnal hypoglycaemia (Pickup and Keen, 2002). However, early studies reported surprisingly little effect on hypoglycaemia, perhaps because of a failure to train patients in the essential related skills of carbohydrate and insulin dose adjustment. More recent work has indicated that CSII can reduce overall rates of hypoglycaemia (Bode et al., 1996; Boland et al., 1999; Kanc et al., 1998), but specific reporting on rates of nocturnal episodes is unusual. Furthermore, few trials have used a randomised design, suggesting that a lower risk might relate to other characteristics of those who use CSII rather than to the technology itself. Thus the amount to which modern pump therapy lowers the risk of nocturnal hypoglycaemia has still be to be established. Nevertheless, in those who have experienced recurrent nocturnal episodes and who have not improved after a trial of insulin analogues, it seems worthwhile undertaking a trial of CSII.
• Nocturnal hypoglycaemia remains an unresolved clinical side-effect of insulin therapy preventing the attainment of strict glycaemic control for many people. It contributes to morbidity, and perhaps mortality, in patients with type 1 diabetes.
• Nocturnal hypoglycaemia is caused chiefly by the limitations of current methods of insulin delivery. The inability of subcutaneous insulin therapy, particularly conventional basal insulins, to maintain low-level stable insulin concentrations overnight, leads both to nocturnal hypoglycaemia and fasting hyperglycaemia.
• In addition to the limitations of insulin delivery, hypoglycaemia is also more common as a result of diminished counterregulatory responses overnight, associated in part with sleep, which has a specific inhibitory effect on physiological defences to hypoglycaemia, and a supine posture which suppresses autonomic responses.
• Nocturnal hypoglycaemia is a particular problem in children, partly as a consequence of the long period of fasting between their evening meal and their breakfast.
• The risk of nocturnal hypoglycaemia is greatest in those who have a bedtime glucose below 7.0 mmol/l. Bedtime snacks can reduce the risk during the early part of the night. Specific foods (uncooked cornstarch or protein rich snacks) reduce the risk of nocturnal hypoglycaemia in some studies.
• Rebound hyperglycaemia (the 'Somogyi phenomenon') is rarely caused by counterregu-latory hormone release provoked by an overnight hypoglycaemic episode, since hormonal secretion is generally suppressed. It is mainly a consequence of waning of circulating plasma insulin levels and should be treated by an adjustment in the timing and type of insulin used rather than a reduction in insulin dose.
• The problem of nocturnal hypoglycaemia may respond to the use of insulin analogues (both rapid and long-acting) or to CSII with an insulin pump. However, its eradication will depend upon new methods of insulin delivery.
Was this article helpful?