Objective. The purpose of this study was to determine whether postprandial administration of the new rapid-acting insulin analog Humalog could effectively reduce glucose excursions in children <5 years old.
Design. Human Regular insulin given before a meal was compared with the same dose of Humalog after a meal of equal carbohydrate content in five toddlers with insulin-dependent (type 1) diabetes mellitus. In addition, the use of Humalog before a meal was compared with Humalog given after a meal of equal carbohydrate content in five other toddlers. The dose of long-acting insulin was not changed during the study period. Blood glucose levels were determined at fasting and at 1, 2, and 4 hours postprandially.
Results. The 2-hour glucose excursions were significantly lower when postprandial Humalog administration was compared with preprandial Human Regular insulin administration. In contrast, glucose excursions were similar when Humalog was taken before or after the meal.
Conclusion. These data show that it is efficacious to give Humalog insulin postprandially in toddlers with type 1 diabetes, allowing increased safety for the young child. The insulin dose can be both matched to the actual food intake and timed to give families increased flexibility and control at mealtime.
Insulin-dependent (type 1) diabetes mellitus is an autoimmune disease that most commonly has its onset in children and young adults. However, preschool children can also be affected, resulting in special problems for the family. One problem is the toddler's refusal to eat after an insulin injection has been given. This can result in severe hypoglycemia and is a serious concern. Feeding problems and control issues may ensue.
The onset of action of Human Regular insulin is between 30 and 180 minutes, and, therefore, the recommended time of administration is least 30 minutes before meals. If it is taken after the meal, as is sometime recommended for toddlers,1 the postprandial blood glucose levels may rise to unacceptably high levels before the onset of insulin activity, mismatching the peak insulin effect and food absorption.
Recently, a new rapid-acting insulin analog Humalog (insulin lispro) has become available. This new insulin has onset of activity at 10 to 15 minutes with peak action occurring between 30 and 90 minutes.2-7 There are no published studies of its use in children, but it may be particularly useful to give after meals in preschool children. The dose then can be reduced (or omitted) if the child eats poorly. The purpose of this study was to compare blood glucose levels after preprandial administration of Human Regular insulin with the blood glucose levels after postprandial administration of Humalog insulin in young children.
MATERIALS AND METHODS
Ten children (Table 1) between the ages of 22 and 58 months (mean, 44 months) and living within a 1-hour drive of the Barbara Davis Center for Childhood Diabetes were studied. The duration of diabetes (Table 1) varied from 5 to 35 months (mean, 20 months). Hemoglobin A1c values (Table 1) were similar for the 10 children.
Nine children were on NPH and one on Lente insulin for their morning dose. Five children were on NPH insulin, one on Lente insulin, and three on Ultralente insulin for their evening dose. One child did not use a long-acting insulin at night (Table 1). There were three boys and seven girls. Children were excluded if they were unable to follow the standardized schedule of 2 study days within 2 weeks or if they were in a period of variable breakfast consumption. There were no other exclusion criteria. Parents signed the consent form approved by the Colorado Multiple Institutional Review Board at the University of Colorado Health Sciences Center, Denver, CO. All children were studied on two separate mornings when healthy and at home to keep the environment as normal as possible.
The children were allowed to eat a breakfast they liked on day 1, and then an identical breakfast was offered on day 2. The breakfasts averaged 50 g of carbohydrate and were ingested within 15 to 20 minutes. The children were assigned to group A or group B based on their current short-acting insulin usage; those using Human Regular insulin were assigned to group A, and those using Humalog were assigned to group B. The five children in group A received Human Regular insulin before the breakfast on day 1 and Humalog insulin immediately after the breakfast on day 2. The timing of the injection of Human Regular insulin depended on the families' usual schedule; if the family usually waited to eat after injecting regular insulin, then this was also done in the study. For the three families that waited, the time between injection and starting breakfast was ∼30 minutes. The five children in group B received Humalog insulin just before the meal on day 1 and Humalog insulin immediately after the meal on day 2. The injections were given by the parent as usual into the same site on both study days. The doses of short- and long-acting insulins were kept the same for both days. All children were able to consume their breakfast within 20 minutes.
Whole blood for glucose levels was obtained by finger prick, using the child's lancet, and determined using the One Touch Profile meter (Johnson and Johnson, Milpitas, CA). Several drops of blood were also collected in a Microtainer-heparinized Plasma Separator Tube (Becton Dickinson and Company, Rutherford, NJ) and transported back to the hospital. The glucose levels were assayed at the hospital using the Yellow Springs Instrument (YSI) 2300 STAT Glucose Analyzer (Yellow Springs Instrument Company, Inc, Yellow Springs, OH) and rechecked with the same One Touch Profile meter as used in the home to confirm the glucose levels. The blood glucose levels determined by the One Touch Profile meter were highly correlated (r = 0.988), with values determined using the YSI. Patients' levels on each day were measured preprandially (fasting) and at 1, 2, and 4 hours after eating their first bite of breakfast. Hemoglobin A1c values were also determined using finger prick whole blood and the DCA 2000 Analyzer (Bayer Corporation, Elkhart, IN). Normal values for this method are 3.4% to 6.2%.
Each child served as his or her own control. Glucose excursions were calculated by subtracting the fasting glucose level from the values at 1, 2, and 4 hours postprandially. The use of glucose excursions makes data comparable from day to day even when fasting blood glucose levels are very different. The use of 2 hour postprandial glucose excursions has been validated as a representative efficacy parameter.7 The glucose levels, as determined by the YSI, were analyzed by the paired Student's t test using the InStat 1.12 statistical program.8
Fasting blood glucose values for the 10 subjects varied from 48 to 292 mg/dL (2.7 to 16.2 mmol/L). The individual glucose excursions for the 5 children who received preprandial Human Regular insulin or postprandial Humalog insulin injections are shown in Fig1 and Table2. The fasting glucose values were similar in the two groups (P > .05) and were not a factor in the glucose excursions. All 5 children had lower 2-hour glucose excursions when the Humalog insulin was given after the meal than when the Human Regular insulin was given before the meal. The mean (±1 SD) glucose excursions for preprandial Human Regular insulin and postprandial Humalog insulin are shown in Fig2. The mean 2-hour postprandial glucose excursions were statistically different (P < .05) for the two groups. Subjects who received Humalog insulin had greater negative glucose excursions compared with those who received Human Regular insulin.
The individual glucose excursions for premeal Humalog insulin compared with postmeal Humalog insulin are shown in Fig3 and Table3. Two of the five children had lower glucose excursions when the Humalog insulin was given postprandially. The fasting glucose levels were not a factor in the individual glucose excursions. The mean 1- , 2- , and 4-hour glucose excursions (Fig4) were similar for the two groups whether the Humalog insulin was given before or after the meal (P > .05).
This is the first study to show that postprandial Humalog insulin administration results in blood glucose values as good as or better than those obtained with the same dose of Human Regular insulin given before the meal. There are no disadvantages to giving Humalog after the meal compared with preprandial Human Regular insulin. The possibility of reducing or omitting the short-acting insulin when the child does not eat offers an important potential advantage.
In a second group of five children, preprandial Humalog insulin was compared with postprandial Humalog insulin. There were no significant differences in the glucose excursions at any period for this group of children. Thus, families with young children using Humalog insulin can safely be flexible with the timing of their injections.
The properties of this new insulin may result in a more physiologic insulin preparation for use in children. Humalog (Lys[B28], Pro[B29], human insulin [recombinant DNA origin]) is an analog of human insulin in which the positions of the amino acids proline (B28) and lysine (B29) in the C-terminal portion of the B-chain are transposed. The amino acid switch results in more rapid dissociation of insulin hexamers. This produces more rapid absorption, an earlier onset of activity (10 to 15 minutes) and earlier peak activity (30 to 90 minutes) than found with Human Regular insulin.2-7The duration of activity was studied by Howey et al2 and is ∼6 hours for Humalog compared with 12 hours for Human Regular insulin. Thus, the activity of Humalog insulin should be adequate to cover the morning snack, particularly because it overlaps with the onset of action of the long-acting insulin. Although not investigated in these patients, our clinical experience shows that it is sometime necessary to add Human Regular insulin to the Humalog insulin to better cover the morning snack and lunch. Clinically, we have noted a delayed onset of Humalog activity when administered with NPH insulin, but not when administered with Lente or Ultralente insulins. As of yet, there is no published research to support these observations, but they may result in changes in the morning Humalog dosage.
Food refusal is a common behavior issue in preschool children. It is a way for the child to exert control over his or her parents, and is generally overcome with patience and understanding. However, it may have serious ramifications in the child with diabetes. Once the child has received an insulin injection, he or she must eat to prevent hypoglycemia. The management of food refusal may now include Humalog insulin given after the meal, with the dose varied to match the amount of food eaten. This allows the parents to remain in control and the phase of food manipulation to pass rapidly.
Hypoglycemia remains the major concern of parents of children with diabetes. Several studies have shown significant reductions in hypoglycemic events when preprandial Humalog insulin was compared with Human Regular insulin.5-7 In our experience, the number of hypoglycemic events was reduced by 50% over 1 year in subjects receiving preprandial Humalog compared with Human Regular insulin.5 The finding of a significant decrease in nocturnal hypoglycemia between the hours of midnight and 6am has also been reported.6,7 The present study did not include adequate numbers of subjects to compare the incidence of hypoglycemia, although two subjects were found to have blood glucose values below 60 mg/dL (<3.25 mmol/L). One subject in group A had a blood glucose level of 54 mg/dL (3.0 mmol/L), and one subject in group B had a blood glucose level of 58 mg/dL (3.2 mmol/L) (both subjects asymptomatic) at 4 hours after postprandial Humalog. No other subjects had blood glucose values <60 mg/dL (3.25 mmol/L) or clinical symptoms of hypoglycemia.
This is the first study to evaluate the use of Humalog insulin after a meal. It is also the first study to describe the use of Humalog insulin in young children. Although not approved by the Food and Drug Administration for use in children, Humalog insulin clearly has special advantages for children that make it just as valuable as for adults. Our clinic has >500 children receiving Humalog insulin, and parents continually note its convenience. Postprandial Humalog insulin administration has now been shown to be effective in achieving lower 2-hour glucose excursions in comparison with preprandial Human Regular insulin. We conclude that postprandial Humalog insulin may increase the safety of insulin administration in young children.
This investigation was supported by Public Health Services, Research Grant 5 M01 RR00051 from the Division of Research Resources and by the Children's Diabetes Foundation.
- Received March 3, 1997.
- Accepted June 10, 1997.
Reprint requests to (H.P.C.) Department of Pediatrics, Barbara Davis Center for Childhood Diabetes, University of Colorado Health Sciences Center, 4200 E 9th Ave, B140, Denver, CO 80262.
- YSI =
- Yellow Springs instrument
- ↵Chase HP. Understanding Insulin-Dependent Diabetes. 8th ed. Denver, CO: Hirschfeld Press; 1995:168
- Howey DC,
- Bowsher RR,
- Brunelle RL,
- Woodworth JR
- DiMarchi RD,
- Chance RE,
- Long HB,
- Shields JE,
- Slieker LJ
- Pfutzer A,
- Kustner E,
- Forst T,
- et al.
- Anderson JH,
- Brunelle RL,
- Koivisto VA,
- et al.
- ↵GraphPad InStat 1.12 for Macintosh. San Diego, CA; GraphPad Software, Inc; 1992
- Copyright © 1997 American Academy of Pediatrics