- Email: [email protected]
Prospective Assessment of Hypoglycemia Symptoms in Children and Adults with Type 1 Diabetes
Can J Diabetes 39 (2015) 26e31
Contents lists available at ScienceDirect
Canadian Journal of Diabetes journal homepage: www.canadianjournalofdiabetes.com
Original Research
Prospective Assessment of Hypoglycemia Symptoms in Children and Adults with Type 1 Diabetes Aditi Amin MPH a, Lorraine Lau PhD a, Susan Crawford MSc b, Alun Edwards MB (FRCPC) c, Danièle Pacaud MD (FRCPC) a, * a
Section of Pediatric Endocrinology and Diabetes, Department of Pediatrics, University of Calgary, Calgary, Alberta, Canada Department of Pediatrics, Alberta Children’s Hospital, Calgary, Alberta, Canada c Section of Endocrinology, Department of Medicine, University of Calgary, Calgary, Alberta, Canada b
a r t i c l e i n f o
a b s t r a c t
Article history: This article was originally published in the August 2014 issue of the Canadian Journal of Diabetes. Received 4 February 2014 Received in revised form 16 May 2014 Accepted 21 May 2014
Purpose: To compare the characteristics of symptoms of hypoglycemia in children and in adults with type 1 diabetes. Methods: Adults with diabetes and parents of children with diabetes who were participants were asked to call a phone system to report episodes of hypoglycemia (presence of symptoms and a blood glucose <4.0 mmol/L). For each episode, blood glucose reading and a scoring of 28 symptoms on a 7-point scale (1 ¼ not present, 7 ¼ very intense) were collected. Results: Sixty six children (49.2% males, mean age ¼ 12.12.4 years, mean age at diagnosis ¼ 7.52.9 years) and 53 adults (41.2% males, mean age 38.714.5 years, mean age at diagnosis ¼ 17.512.9 years) with type 1 diabetes participated. The most common symptoms in adults were hunger, sweating, trembling and weakness. The most common symptoms in children were weakness, trembling and hunger. The 2 most discriminating variables between children and adults were sleepiness and tiredness, which were more common in children (p<0.01). In a comparative factor analysis, 3 factors emerged: factor 1, autonomic and neuroglycopenic; factor 2, behavioural; and factor 3, general malaise. Factors 2 and 3 were significantly more common or intense in children than in adults; MANOVA: F(1, 113) ¼ 6.72, p<0.05 and F(1, 113) ¼ 4.64, p<0.05, respectively. Conclusions: Symptoms relating to behaviour and general malaise are more common in children than in adults with type 1 diabetes. The results of this study may assist providers in educating caregivers of children and patients with diabetes how to better recognize episodes of hypoglycemia. Ó 2015 Canadian Diabetes Association
Keywords: adults children and adolescents hypoglycemia symptoms type 1 diabetes
r é s u m é Mots clés : adultes enfants et adolescents symptômes de l’hypoglycémie diabète de type 1
Objet : Comparer les caractéristiques des symptômes de l’hypoglycémie chez les enfants et les adultes atteints du diabète de type 1. Méthodes : Les participants, des adultes atteints de diabète et les parents d’enfants diabétiques, étaient invités à appeler un système téléphonique pour rapporter les épisodes d’hypoglycémie (présence de symptômes et glycémie < 4,0 mmol/l). À chacun des épisodes, la lecture de la glycémie et l’évaluation de 28 symptômes sur une échelle en 7 points (1 ¼ non présent, 7 ¼ très intense) étaient recueillies. Résultats : Soixante-six (66) enfants (49,2 % de garçons, âge moyen de 12,1 2,4 ans, âge moyen au diagnostic de 7,5 2,9 ans) et 53 adultes (41,2 % d’hommes, âge moyen de 38,7 14,5 ans, âge moyen au diagnostic de 17,5 12,9 ans) atteints du diabète de type 1 ont participé. Les symptômes les plus fréquents chez les adultes étaient la faim, la sudation, le tremblement et la faiblesse. Les symptômes les plus fréquents chez les enfants étaient la faiblesse, le tremblement et la faim. Les 2 variables les plus discriminantes entre les enfants et les adultes étaient la somnolence et la fatigue, qui étaient plus fréquentes chez les enfants (p < 0,01). Dans une analyse factorielle comparative, 3 facteurs sont ressortis : facteur 1, autonome et neuroglycopénique; facteur 2, comportemental; facteur 3, malaise général. Les facteurs 2 et 3 étaient significativement plus fréquents ou intenses chez les enfants que chez les adultes; MANOVA : F(1, 113) ¼ 6,72, p < 0,05 et F(1, 113) ¼ 4,64, p0,05, respectivement.
* Address for correspondence: Danièle Pacaud, MD, Alberta Children’s Hospital, 2888 Shaganappi Trail NW, Calgary, Alberta T3B 6A8, Canada. E-mail address: [email protected] 1499-2671/$ e see front matter Ó 2015 Canadian Diabetes Association http://dx.doi.org/10.1016/j.jcjd.2015.09.086
A. Amin et al. / Can J Diabetes 39 (2015) 26e31
27
Conclusions : Les symptômes relatifs au comportement et au malaise général sont plus fréquents chez les enfants que chez les adultes atteints du diabète de type 1. Les résultats de cette étude peuvent aider à sensibiliser les aidants et les soignants des enfants et des patients souffrant de diabète à mieux reconnaître les épisodes d’hypoglycémie. Ó 2015 Canadian Diabetes Association
Introduction Hypoglycemia is the most common complication in the treatment of type 1 diabetes mellitus (1,2). The Diabetes Complication and Control Trial (DCCT) has demonstrated the importance of tighter metabolic control in preventing long-term complications of type 1 diabetes (3). However, intensified diabetes management is also associated with increased frequency of hypoglycemic events, which, particularly if severe, becomes a barrier to tighter metabolic control (1,4). Hypoglycemia is classically defined and diagnosed by a combination of symptoms and biochemical criteria known as the Whipple triad; symptoms typical of hypoglycemia are confirmed by low plasma glucose levels and are relieved by carbohydrate intake. The difficulty is in the definition of symptoms as typical. The spectrum of symptoms of hypoglycemia is broad and may vary with time or may fluctuate, often as an alteration in neurohormonal responses after preceding episodes of hypoglycemia (4). Symptoms of hypoglycemia, when classically defined and confirmed through principal component analysis, have been divided into 3 categories: autonomic, neuroglycopenic and general malaise. Autonomic symptoms include sweating, palpitations, shaking and hunger. Neuroglycopenic symptoms include confusion, drowsiness, speech difficulty and odd behaviour. Finally, general malaise includes nausea and headaches (4e7). Deary and his group studied hypoglycemic symptoms in adults with type 1 diabetes using factor analysis and have shown very good segregation of symptoms into the autonomic or neuroglycopenic categories. These autonomic and neuroglycopenic symptom groups were associated with different glycemic thresholds, depending on the severity of the episode (7). The results of the DCCT showed a 3-fold increased relative risk for severe hypoglycemia (3) with intensive diabetes management. The DCCT included a subgroup of 195 adolescents (13 to 18 years of age), which allowed comparison of this age group with the adult participants. In both the conventional and the intensified diabetes management groups, the adolescents had significantly higher rates of hypoglycemia compared to the adults (8,9). Although some studies (10e12) suggest differing presentation of symptoms of hypoglycemia in children and in adults, to our knowledge, there is no published study specifically designed to compare symptomatology in adults and children. This may be an important distinction because children are at increased risk for hypoglycemia (13,14) and because the ability of children and their parents to detect symptoms of hypoglycemia has been reported as being poor (15). An understanding of these differences in symptomatology is needed to better educate caregivers and patients with type 1 diabetes and to improve their abilities to recognize and treat hypoglycemia. Therefore, we conducted this study to report and compare the symptoms of hypoglycemia in children and adults. Methods This study was approved by the Conjoint Health Research Ethics Board at the University of Calgary. Subjects were recruited from patients aged 6 to 18 years of age who had type 1 diabetes and were followed at the Alberta Children’s Hospital and from adult subjects
with type 1 diabetes attending the Calgary Health Region Diabetes, Hypertension and Cholesterol Centre. Individuals were excluded based on the following criteria: inadequate proficiency in English; physical or intellectual barriers to completion of study; use of medication interfering with ability to perceive hypoglycemia symptoms (e.g. beta-blockers, oral corticosteroids, etc); and other concomitant medical problems requiring ongoing treatment other than controlled asthma, celiac disease, hypothyroidism, hypertension or hyperlipidemia. Once informed consent had been obtained, the subjects and 1 parent were asked to complete a validated 28-item questionnaire on hypoglycemia symptoms (16) at baseline (retrospective symptom report). Children completed the survey with a research assistant, and the parents of pediatric subjects completed their versions in a separate room. Adult participants completed the questionnaire on their own during the initial meeting with the research assistant. For each symptom, a score from 1 to 7 could be assigned (1 ¼ symptom not present to 7 ¼ very intense). Symptoms were divided into 3 main groups: autonomic, neuroglycopenic and general malaise. Thereafter, subjects were asked to call a specialized automated phone system to report each episode of hypoglycemia (symptoms of hypoglycemia and blood glucose <4.0 mmol/L on monitoring) for a 4-week period following their enrolment. Date, time, level of blood glucose and rating for each of the 28 items on the hypoglycemia questionnaire were collected from each call. For younger children, the parents were instructed to have the child rate each symptom, even if the parent was phoning in for them. To help with interpretation of the results, the following information was obtained from the patient or the patient’s medical chart: date of birth, date of diagnosis, gender, past history of severe hypoglycemia and ability to tell whether having hypoglycemia (single question with a 6-point Likert scale: 1 ¼ always to 6 ¼ never). The patient was also asked to obtain an A1C result if this result had not been obtained in a 3-month period prior to enrolment. At the end of the study, subjects were asked to fax or mail a copy of their logbooks for 4 weeks of the study. For the 66 subjects for whom a logbook was available (42 children and 24 adults), 31 had good concordance between the reported frequency of hypoglycemia, whereas 35 had discrepancies of more than 1 episode between the logbook and the phone system; 26 had more episodes reported by phone than in the logbook. Therefore, compliance with the system was good or very good in 86% (57/66) of subjects. Data Analysis Comparisons of retrospective and prospective data (between adults and children with type 1 diabetes) for each of the 28 individual symptoms were conducted using analysis of variance (ANOVA) to compare adults to children and repeated ANOVA measures to compare retrospective to prospective data within each patient group. A discriminant function analysis was conducted using all 28 symptoms for children and for adults so as to determine which combination of the symptoms best discriminated between adults and children. The percentage of adults correctly classified as adults and the percentage of children correctly classified as children were reported. In order to avoid interference by varying numbers of
28
A. Amin et al. / Can J Diabetes 39 (2015) 26e31
Table 1 Comparative factor analysis
Factor 1 Autonomic and neuroglycopenic symptoms
Factor 2 Behavioural symptoms
Factor 3 General malaise symptoms
Table 2 Sample characteristics Retrospective recall
Prospective report
Variable
Sweating Warmth Confusion Trembling Restlessness Slurred speech Weakness Pounding heart Yawning Tingling lips Daydreaming Blurred vision Argumentativeness, defiance Irritability Aggressiveness Uncoordination Odd behaviour Feeling awful Dizziness Sleepiness Hunger Tummy pain Tiredness Feeling sick Headache Tearfulness
Sweating Warmth Confusion Trembling Restlessness Slurred speech Weakness Pounding heart Uncoordination Feeling sick Odd behaviour Dizziness Argumentativeness Defiance Irritablility Aggressiveness Headache Tearfulness Sleepiness Hunger Tiredness Tummy pain Feeling awful Yawning
Children (n¼66)
Adults (n¼53)
Proportion of males (%) Age (years) at assessment (mean SD) Age (years) at diagnosis (mean SD) Hypothyroid (%) Hypertension (%) Nephropathy (%) Retinopathy (%) Neuropathy (%) Number of hypoglycemic episodes in 4 weeks >2 hypoglycemic episodes/week Lifelong history of severe hypoglycemia Ability to tell if has hypoglycemia A1C (%, normal <6.1%) Total daily dose (units/kg, mean SD) Insulin type (%) Rapid-acting analogue Long-acting analogue Continued insulin infusion pump
49.2% males 12.12.4
41.2% males 38.714.5
NS p<0.001
7.52.9
17.512.9
p<0.001
1.6% 0% 0% 0% 0% 8.06.5
20% 14% 8% 34% 20% 10.511.0
p<0.001 p<0.01 p<0.05 p<0.001 p<0.001 NS
35.4% 59.4%
41.5% 66.7%
NS NS
92% 8.41.2 0.990.37
86% 7.81.3 0.660.26
NS p<0.05 p<0.001
74.2% 0% 12.1%
62.3% 15.1% 9.4%
p<0.001
episodes in each individual, this analysis was done using the first prospectively reported episode only. A comparative factor analysis was used to determine whether the 28 symptoms co-segregated into the 3 groups observed by Deary et al (7) in both adults and children. The equality of the factor structures across groups (i.e. those obtained for adults and those obtained for children) was determined using multisample confirmatory factor analysis with factor loadings >.3 as the cut-off. The same strategy was used to compare retrospective and prospective symptom reports. When conducting comparative factor analysis from both retrospective recall and prospective report of symptoms, 3 similar but nonidentical factors emerged, as seen in Table 1. Results Participants included 66 children (49.2% males; mean age ¼ 12.1 2.4 years; mean age at diagnosis ¼ 7.52.9 years) through parent report and 53 adults (41.2% males; mean age ¼ 38.714.5 years; mean age at diagnosis ¼ 17.512.9 years) with type 1 diabetes. Of the children, 53 (80%) were younger than 13 years of age, and 13 (20%) were 13 years of age or older. Additional characteristics of the sample are shown in Table 2. The groups differed in the duration of disease and the presence of complications. Children and adults did not differ in their abilities to know when their blood sugar levels were low or in their A1C levels, frequency of hypoglycemia or history of severe hypoglycemia. No participants reported having severe hypoglycemia during the month of observation. Children, however, had significantly higher scores for being tearful (p<0.01), being aggressive (p<0.05), having abdominal pain (p<0.01), being hungry (p<0.01) and being dizzy (p<0.05) than did adults. Adults had significantly higher scores for sweating (p<0.01), daydreaming (p<0.001), and having tingling lips (p<0.01) as compared to children. Analysis comparing childrens’ and adults’ retrospective recall of symptoms As shown in Figure 1, reported intensity of symptoms was compared between the retrospective reports and the prospective
NS, Not significant.
reports (the first episodes called in). Results showed that for adults, significantly higher scores emerged in retrospective data for all symptoms except 3: tearfulness, headache and abdominal pain. A similar pattern emerged when children were analyzed separately: significantly higher scores for retrospective data for all 28 symptoms except blurred vision and tummy pain. When conducting comparative factor analysis from the retrospective recall of symptoms, adults and children differed significantly on all 3 factors: adults reported higher scores on factor 1 (F[1,173] ¼ 15.67, p<0.001) whereas children had higher scores on factor 2 (F[1,173] ¼ 4.58, p<0.05) and factor 3 (F[1,173] ¼ 35.92, p<0.001). When the factor analysis was repeated after separating children into 2 age groups (<13 years and 13 years), only 2 distinct factors emerged (37.8% of variance explained) for the younger group, whereas 3 factors emerged (41.5% of variance explained) in the older group (data not shown). Analysis comparing childrens’ and parents’ retrospective recall of symptoms For each symptom, the rating given by the child was correlated with the rating given by the parents. Parents’ reports were significantly correlated with childrens’ reports for every symptom, with the exception of sleepiness, nightmares, slurred speech, blurry vision, tummy pain, warmth, restlessness, daydreaming, pounding heart, tingling lips and feeling awful (data not shown). Analyses using first prospectively reported episode The most common symptoms for the first episode called in for the study of adults were hunger, sweating, trembling and weakness. For children, weakness, trembling and hunger were more common. Figure 2 shows the comparison between children and adults of the symptom scores on the first prospective episode reported. The 2 most discriminating symptoms for adults vs. children for the first episode called in for the study were sleepiness and tiredness (overall model for discriminant function analysis, c2 (28, N ¼ 119) ¼ 58.27, p<0.01). Overall, 81.1% of adults and 82.5% of
A. Amin et al. / Can J Diabetes 39 (2015) 26e31
29
A
B
Figure 1. Retrospective and Prospective Symptom Reporting. Panel (A), adults; panel (B), children. The intensity of symptoms was significantly different between the retrospective recall and the prospective report, except for those marked with n/s (nonsignificant; p>0.05).
children were correctly classified as adults (or children). Being tired and feeling sleepiness were significantly more intense and more common in children (p<0.01). In a comparative factor analysis using all 28 symptoms for the first episode called in with oblique rotation, 3 factors emerged that explained 42.7% of the variance (see Table 1). As observed in the retrospective comparison, the prospective reports showed that hypoglycemic symptoms in children and adults differed significantly in terms of behavioural symptoms (factor 2) and general malaise symptoms (factor 3) but not in terms of autonomic and neuroglycopenic symptoms (factor 1). Behavioural and general malaise symptoms were significantly more common and more intense in children than in adults (F[1, 113]¼ 6.72, p<0.05 and F [1, 113] ¼ 4.64, p<0.05, respectively). The correlations between blood glucose levels and each of the 3 factor scores were not significant for adults or for children. Thus, in these 2 groups, the severity of symptoms was not linked to the actual blood glucose levels. Analyses using number of hypoglycemic episodes When the sample was stratified based on the number of hypoglycemic episodes experienced per week (i.e. <1/week, 1 to 3/week, >3/week), the largest proportion of adults (39%) and the majority of children (52%) had 1 to 3 episodes per week. The association between the number of episodes and whether the subject was a child or an adult was not significant. As shown in Figure 3, as expected, there was an inverse relationship between A1C levels and the number of hypoglycemic episodes per week in both
children and adults. The correlation between A1C levels and number of hypoglycemic episodes per week was significant for adults (r¼.378, p¼0.023), and the same pattern, albeit a trend, emerged for children (r¼.264, p¼0.084). The correlation did not differ significantly between adults and children (p¼0.589). For adults, there was a significant correlation between more hypoglycemic episodes per week and lower ratings of hunger (r¼.422, p¼0.003) but higher ratings for feelings of sickness (r¼.297, p¼0.040) and tearfulness (r¼.318, p¼0.028). For children, there was a significant correlation between more hypoglycemic episodes per week and better ability to tell whether having hypoglycemia (r¼.314, p¼0.014). Discussion and Conclusions We present the findings of this study believing it to be unique in comparing directly the differences in hypoglycemia symptomatology between children and adults with type 1 diabetes. Furthermore, only a few studies (15,17) have compared patients’ recall of hypoglycemia with actual events. This study also directly compared patients’ reported symptoms both retrospectively (i.e. recall) and prospectively (in-the-moment reporting) in contrast to past studies that examined the accuracy of reporting of severe hypoglycemic events rather than actual symptoms. Both children and adults had higher symptom scores by recollection than were achieved when they reported events during the study. This finding suggests that when hypoglycemic, patients with type 1 diabetes may have fewer symptoms than they think they have, with the potential that they are inclined to delay treatment.
30
A. Amin et al. / Can J Diabetes 39 (2015) 26e31
Figure 2. Comparison of symptoms scores on the first prospective event between adults and children. n/s: nonsignificant difference between children and adults.
It may be, however, that enrolment in the study, with detailed enquiry about their symptoms and knowing they had to report each episode for a month, they became sensitized and identified episodes earlier. Even if this latter explanation holds true, it supports the benefit of hypoglycemia awareness training to minimize the risk for severe hypoglycemic episodes (18,19). As such, providers should strive to work with patients with type 1 diabetes and their caregivers to identify the less intense symptoms that can serve as warnings for hypoglycemic episodes and enable patients and caregivers to treat hypoglycemic episodes early and reduce the risk for severe hypoglycemia. The results of this study also verify that presenting symptoms of hypoglycemia in children differ from those in adults. Tupola et al (20), using diaries of symptoms of hypoglycemia over a 3-month period in 161 children with type 1 diabetes, found that during mild episodes, the predominant symptoms were nonspecific in 41% of episodes, neuroglycopenic in 20% of episodes and autonomic in 39%
Figure 3. A1C for children and adults by number of hypoglycemic episodes per week. Note that only 44 of 66 children and 36 of 53 adults had A1C levels that were available within 3 months of the study.
of episodes. In children younger than 6 years, only 34% of episodes had predominantly autonomic symptoms compared with 57% of episodes in adolescents 15 years of age and older (p¼0.01). The most common symptoms reported in that study were weakness, tremor, hunger and drowsiness. Similarly, in our study, children more commonly reported weakness, trembling and hunger as symptoms of hypoglycemia. The symptoms of being tired and sleepy were also found to be significantly more intense and common in children than in the adults in our study. McCrimmon et al (11) interviewed 100 parents and 43 children with diabetes. They found that both autonomic and neuroglycopenic symptoms seem to be present at the same time in children. This was in contrast to adults, in whom autonomic and neuroglycopenic symptoms were well identified as being separated or hierarchical events, depending on the severity of the hypoglycemia. Our analysis failed to replicate this finding in adults; however, that may be due to the method used for glucose measurement (home glucose meter measurements vs. laboratory measurements; see discussion below in the Limitations section). McCrimmon et al (11) also reported that both parents and children identified behavioural changes as frequent and early signs of hypoglycemia. The results of our study corroborate the findings of McCrimmon et al, in that behavioural and general malaise symptoms were significantly more common and intense in children than in adults. One could postulate that the reported increase in behavioural symptoms in children is the result of an observer bias (symptoms of children are reported by parents; adults self-report). However, when we compared the symptoms reported by the children with those reported by their parents, we found a good correlation between children’s and parents’ reports for most symptoms, including behavioural symptoms. This difference between adults and adolescents may be explained in part by the differing responses to hypoglycemia seen in children compared to adults. Jones et al (10) studied the counterregulatory hormone and symptomatic responses to hypoglycemia in 13 poorly controlled children with type 1 diabetes and compared the responses to those of 16 children and 19 adults without diabetes. In children with diabetes, hypoglycemic symptoms were
A. Amin et al. / Can J Diabetes 39 (2015) 26e31
identified at glucose levels of 4.2 mmol/L compared to 3.6 mmol/L in children without diabetes and less than 3.0 mmol/L in adults without diabetes. While experiencing low blood glucose levels, the 2 groups of children had significantly higher symptom scores than did the adult group. The epinephrine and growth hormone responses paralleled the symptom scores, with more rapid and greater responses in children with diabetes. Both groups of children had significantly higher responses compared to the adult group at the end of the study. Ross et al also found that although symptom responses were similar in prepubertal, midpubertal and postpubertal children with type 1 diabetes, prepubertal children had lower blood glucose levels when autonomic reaction was observed (21). They also noted that the sweating and pancreatic polypeptide responses to hypoglycemia present in adults were seen only in postpubertal children. Limitations of this study include the nature of children’s maturation, which may not allow prepubertal children to comprehend and express some symptoms of hypoglycemia. The real-life setting of this study also limited our ability to have accurate laboratory blood glucose levels measured, which may explain the lack of hierarchy or progression of symptoms with decreasing blood glucose levels that may be observed in laboratory settings. Home glucose monitors are known be less accurate in lower glucose ranges than in normal or hyperglycemic ranges. Another limitation of this study is the difference between groups in duration of diabetes and presence of microvascular complications. However, if this difference had an effect, it might reduce the frequency or severity of autonomic symptoms in the adult group and would bias against finding a difference between the children and the adults. Finally, the number of missing A1C results limits the generalization of our results associated with this parameter. We also recognize that the nature of hypoglycemic symptoms can vary or fluctuate in individual patients with diabetes (perhaps related to altered neurohormonal responses after preceding episodes of hypoglycemia). This is, in part, the reason for our design of examining the patient recall (retrospective) of symptoms as well as the current state (prospective). Overall, this study corroborates findings that hypoglycemia symptomatology differs significantly between adults and children and notes the importance of improved recognition and treatment of hypoglycemia in both of these groups. Healthcare providers should be aware of the differing presentations of hypoglycemia between adults and children and should consider these differences when educating patients and their caregivers in developing individualized care plans. Providers should also encourage patients and their caregivers to perform regular blood glucose testing and monitoring to prevent, detect and enable early treatment of hypoglycemia. Additionally, providers should work with children with type 1 diabetes to ensure that they are aware of how their hypoglycemia symptoms will evolve as they transition between childhood and adulthood so they can move away from dependency on their caregivers to recognize hypoglycemia symptoms and reach self-recognition of hypoglycemia symptoms.
31
Acknowledgements We are grateful to the families and patients who agreed to participate in the study. This study was supported by a grant from the Stewart Fund for Diabetes Research. All authors had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. References 1. Davis E, Keating B, Byrne G, Russell M, Jones T. Impact of improved glycaemic control on rates of hypoglycaemia in insulin-dependent diabetes mellitus. Arch Dis Child 1998;78:111e5. 2. Clayton D, Woo V, Yale J. Hypoglycemia. Can J Diabetes 2013;37(Suppl 1): S69e71. 3. Diabetes Control and Complications Trial Research Group. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med 1993; 329:977e86. 4. Cryer PE. The barrier of hypoglycemia in diabetes. Diabetes 2008;57: 3169e76. 5. Hepburn DA, Deary IJ, Frier BM, et al. Symptoms of acute insulin-induced hypoglycemia in humans with and without IDDM factor-analysis approach. Diabetes Care 1991;14:949e57. 6. Hepburn DA, Dary IJ, Frier BM. Classification of symptoms of hypoglycemia in insulin-treated diabetic patients using factor analysis: Relationship to hypoglycemia unawareness. Diabet Med 1992;9:70e5. 7. Deary IJ, Hepburn DA, MacLeod KM, Brier BM. Partitioning the symptoms of hypoglycemia using multi-sample confirmatory factor analysis. Diabetologia 1993;36:771e7. 8. Diabetes Control and Complications Trial Research Group. Hypoglycemia in the diabetes control and complications trial. Diabetes 1997;46:271e86. 9. Diabetes Control and Complications Trial Research Group. Effect of intensive diabetes treatment on the development and progression of long-term complications in adolescents with insulin-dependent diabetes mellitus: Diabetes Control and Complications Trial. J Pediatr 1994;125:177e88. 10. Jones T, Boulware S, Kraemer D, et al. Independent effects of youth and poor diabetes control on responses to hypoglycemia in children. Diabetes 1991;40: 358e63. 11. McCrimmon R, Gold A, Deary I. Symptoms of hypoglycemia in children with IDDM. Diabetes Care 1995;18:858e61. 12. Graveling AJ, Noyes KJ, Allerhand MH, et al. Prevalence of impaired awareness of hypoglycemia and identification of predictive symptoms in children and adolescents with type 1 diabetes. Pediatr Diabetes 2014;15:206e13. 13. Pacaud D. Hypoglycemia: the Achilles heel of the treatment of children with type 1 diabetes. Can J Diabetes 2002;26:215e22. 14. Wherrett D, Huot C, Pacaud D. Type 1 diabetes in children and adolescents. Can J Diabetes 2013;37(Suppl 1):S153e62. 15. Gonder-Frederick LJ, Zrebiec A, Bauchowitz J, et al. Detection of hypoglycemia by children with type 1 diabetes 6 to 11 years of age and their parents: A field study. Pediatrics 2008;121:e489e95. 16. Ross LA, McCrimmon RJ, Frier BJ, et al. Hypoglycaemic symptoms reported by children with type 1 diabetes mellitus and by their parents. Diabet Med 1998; 15:836e43. 17. Davis E, Keating B, Byrne G, et al. Hypoglycemia: Incidence and clinical predictors in a large population-based sample of children and adolescents with IDDM. Diabetes Care 1997;20:22e5. 18. Cox DJ, Gonder-Frederick L, Polonsky W, et al. Blood glucose awareness training (BGAT-2): Long-term benefits. Diabetes Care 2001;24:637e42. 19. Broers S, le Cessie S, van Vliet KP, et al. Blood glucose awareness training in Dutch type 1 diabetes patients: Short-term evaluation of individual and group training. Diabet Med 2002;19:157e61. 20. Tupola S, Rajantie J. Documented symptomatic hypoglycaemia in children and adolescents using multiple daily insulin injection therapy. Diabet Med 1998; 15:492e6. 21. Ross LA, Warren RE, Kelnar CJ, Frier BM. Pubertal stage and hypoglycaemia counterregulation in type 1 diabetes. Arch Dis Child 2005;90:190e4.
Contents lists available at ScienceDirect
Canadian Journal of Diabetes journal homepage: www.canadianjournalofdiabetes.com
Original Research
Prospective Assessment of Hypoglycemia Symptoms in Children and Adults with Type 1 Diabetes Aditi Amin MPH a, Lorraine Lau PhD a, Susan Crawford MSc b, Alun Edwards MB (FRCPC) c, Danièle Pacaud MD (FRCPC) a, * a
Section of Pediatric Endocrinology and Diabetes, Department of Pediatrics, University of Calgary, Calgary, Alberta, Canada Department of Pediatrics, Alberta Children’s Hospital, Calgary, Alberta, Canada c Section of Endocrinology, Department of Medicine, University of Calgary, Calgary, Alberta, Canada b
a r t i c l e i n f o
a b s t r a c t
Article history: This article was originally published in the August 2014 issue of the Canadian Journal of Diabetes. Received 4 February 2014 Received in revised form 16 May 2014 Accepted 21 May 2014
Purpose: To compare the characteristics of symptoms of hypoglycemia in children and in adults with type 1 diabetes. Methods: Adults with diabetes and parents of children with diabetes who were participants were asked to call a phone system to report episodes of hypoglycemia (presence of symptoms and a blood glucose <4.0 mmol/L). For each episode, blood glucose reading and a scoring of 28 symptoms on a 7-point scale (1 ¼ not present, 7 ¼ very intense) were collected. Results: Sixty six children (49.2% males, mean age ¼ 12.12.4 years, mean age at diagnosis ¼ 7.52.9 years) and 53 adults (41.2% males, mean age 38.714.5 years, mean age at diagnosis ¼ 17.512.9 years) with type 1 diabetes participated. The most common symptoms in adults were hunger, sweating, trembling and weakness. The most common symptoms in children were weakness, trembling and hunger. The 2 most discriminating variables between children and adults were sleepiness and tiredness, which were more common in children (p<0.01). In a comparative factor analysis, 3 factors emerged: factor 1, autonomic and neuroglycopenic; factor 2, behavioural; and factor 3, general malaise. Factors 2 and 3 were significantly more common or intense in children than in adults; MANOVA: F(1, 113) ¼ 6.72, p<0.05 and F(1, 113) ¼ 4.64, p<0.05, respectively. Conclusions: Symptoms relating to behaviour and general malaise are more common in children than in adults with type 1 diabetes. The results of this study may assist providers in educating caregivers of children and patients with diabetes how to better recognize episodes of hypoglycemia. Ó 2015 Canadian Diabetes Association
Keywords: adults children and adolescents hypoglycemia symptoms type 1 diabetes
r é s u m é Mots clés : adultes enfants et adolescents symptômes de l’hypoglycémie diabète de type 1
Objet : Comparer les caractéristiques des symptômes de l’hypoglycémie chez les enfants et les adultes atteints du diabète de type 1. Méthodes : Les participants, des adultes atteints de diabète et les parents d’enfants diabétiques, étaient invités à appeler un système téléphonique pour rapporter les épisodes d’hypoglycémie (présence de symptômes et glycémie < 4,0 mmol/l). À chacun des épisodes, la lecture de la glycémie et l’évaluation de 28 symptômes sur une échelle en 7 points (1 ¼ non présent, 7 ¼ très intense) étaient recueillies. Résultats : Soixante-six (66) enfants (49,2 % de garçons, âge moyen de 12,1 2,4 ans, âge moyen au diagnostic de 7,5 2,9 ans) et 53 adultes (41,2 % d’hommes, âge moyen de 38,7 14,5 ans, âge moyen au diagnostic de 17,5 12,9 ans) atteints du diabète de type 1 ont participé. Les symptômes les plus fréquents chez les adultes étaient la faim, la sudation, le tremblement et la faiblesse. Les symptômes les plus fréquents chez les enfants étaient la faiblesse, le tremblement et la faim. Les 2 variables les plus discriminantes entre les enfants et les adultes étaient la somnolence et la fatigue, qui étaient plus fréquentes chez les enfants (p < 0,01). Dans une analyse factorielle comparative, 3 facteurs sont ressortis : facteur 1, autonome et neuroglycopénique; facteur 2, comportemental; facteur 3, malaise général. Les facteurs 2 et 3 étaient significativement plus fréquents ou intenses chez les enfants que chez les adultes; MANOVA : F(1, 113) ¼ 6,72, p < 0,05 et F(1, 113) ¼ 4,64, p0,05, respectivement.
* Address for correspondence: Danièle Pacaud, MD, Alberta Children’s Hospital, 2888 Shaganappi Trail NW, Calgary, Alberta T3B 6A8, Canada. E-mail address: [email protected] 1499-2671/$ e see front matter Ó 2015 Canadian Diabetes Association http://dx.doi.org/10.1016/j.jcjd.2015.09.086
A. Amin et al. / Can J Diabetes 39 (2015) 26e31
27
Conclusions : Les symptômes relatifs au comportement et au malaise général sont plus fréquents chez les enfants que chez les adultes atteints du diabète de type 1. Les résultats de cette étude peuvent aider à sensibiliser les aidants et les soignants des enfants et des patients souffrant de diabète à mieux reconnaître les épisodes d’hypoglycémie. Ó 2015 Canadian Diabetes Association
Introduction Hypoglycemia is the most common complication in the treatment of type 1 diabetes mellitus (1,2). The Diabetes Complication and Control Trial (DCCT) has demonstrated the importance of tighter metabolic control in preventing long-term complications of type 1 diabetes (3). However, intensified diabetes management is also associated with increased frequency of hypoglycemic events, which, particularly if severe, becomes a barrier to tighter metabolic control (1,4). Hypoglycemia is classically defined and diagnosed by a combination of symptoms and biochemical criteria known as the Whipple triad; symptoms typical of hypoglycemia are confirmed by low plasma glucose levels and are relieved by carbohydrate intake. The difficulty is in the definition of symptoms as typical. The spectrum of symptoms of hypoglycemia is broad and may vary with time or may fluctuate, often as an alteration in neurohormonal responses after preceding episodes of hypoglycemia (4). Symptoms of hypoglycemia, when classically defined and confirmed through principal component analysis, have been divided into 3 categories: autonomic, neuroglycopenic and general malaise. Autonomic symptoms include sweating, palpitations, shaking and hunger. Neuroglycopenic symptoms include confusion, drowsiness, speech difficulty and odd behaviour. Finally, general malaise includes nausea and headaches (4e7). Deary and his group studied hypoglycemic symptoms in adults with type 1 diabetes using factor analysis and have shown very good segregation of symptoms into the autonomic or neuroglycopenic categories. These autonomic and neuroglycopenic symptom groups were associated with different glycemic thresholds, depending on the severity of the episode (7). The results of the DCCT showed a 3-fold increased relative risk for severe hypoglycemia (3) with intensive diabetes management. The DCCT included a subgroup of 195 adolescents (13 to 18 years of age), which allowed comparison of this age group with the adult participants. In both the conventional and the intensified diabetes management groups, the adolescents had significantly higher rates of hypoglycemia compared to the adults (8,9). Although some studies (10e12) suggest differing presentation of symptoms of hypoglycemia in children and in adults, to our knowledge, there is no published study specifically designed to compare symptomatology in adults and children. This may be an important distinction because children are at increased risk for hypoglycemia (13,14) and because the ability of children and their parents to detect symptoms of hypoglycemia has been reported as being poor (15). An understanding of these differences in symptomatology is needed to better educate caregivers and patients with type 1 diabetes and to improve their abilities to recognize and treat hypoglycemia. Therefore, we conducted this study to report and compare the symptoms of hypoglycemia in children and adults. Methods This study was approved by the Conjoint Health Research Ethics Board at the University of Calgary. Subjects were recruited from patients aged 6 to 18 years of age who had type 1 diabetes and were followed at the Alberta Children’s Hospital and from adult subjects
with type 1 diabetes attending the Calgary Health Region Diabetes, Hypertension and Cholesterol Centre. Individuals were excluded based on the following criteria: inadequate proficiency in English; physical or intellectual barriers to completion of study; use of medication interfering with ability to perceive hypoglycemia symptoms (e.g. beta-blockers, oral corticosteroids, etc); and other concomitant medical problems requiring ongoing treatment other than controlled asthma, celiac disease, hypothyroidism, hypertension or hyperlipidemia. Once informed consent had been obtained, the subjects and 1 parent were asked to complete a validated 28-item questionnaire on hypoglycemia symptoms (16) at baseline (retrospective symptom report). Children completed the survey with a research assistant, and the parents of pediatric subjects completed their versions in a separate room. Adult participants completed the questionnaire on their own during the initial meeting with the research assistant. For each symptom, a score from 1 to 7 could be assigned (1 ¼ symptom not present to 7 ¼ very intense). Symptoms were divided into 3 main groups: autonomic, neuroglycopenic and general malaise. Thereafter, subjects were asked to call a specialized automated phone system to report each episode of hypoglycemia (symptoms of hypoglycemia and blood glucose <4.0 mmol/L on monitoring) for a 4-week period following their enrolment. Date, time, level of blood glucose and rating for each of the 28 items on the hypoglycemia questionnaire were collected from each call. For younger children, the parents were instructed to have the child rate each symptom, even if the parent was phoning in for them. To help with interpretation of the results, the following information was obtained from the patient or the patient’s medical chart: date of birth, date of diagnosis, gender, past history of severe hypoglycemia and ability to tell whether having hypoglycemia (single question with a 6-point Likert scale: 1 ¼ always to 6 ¼ never). The patient was also asked to obtain an A1C result if this result had not been obtained in a 3-month period prior to enrolment. At the end of the study, subjects were asked to fax or mail a copy of their logbooks for 4 weeks of the study. For the 66 subjects for whom a logbook was available (42 children and 24 adults), 31 had good concordance between the reported frequency of hypoglycemia, whereas 35 had discrepancies of more than 1 episode between the logbook and the phone system; 26 had more episodes reported by phone than in the logbook. Therefore, compliance with the system was good or very good in 86% (57/66) of subjects. Data Analysis Comparisons of retrospective and prospective data (between adults and children with type 1 diabetes) for each of the 28 individual symptoms were conducted using analysis of variance (ANOVA) to compare adults to children and repeated ANOVA measures to compare retrospective to prospective data within each patient group. A discriminant function analysis was conducted using all 28 symptoms for children and for adults so as to determine which combination of the symptoms best discriminated between adults and children. The percentage of adults correctly classified as adults and the percentage of children correctly classified as children were reported. In order to avoid interference by varying numbers of
28
A. Amin et al. / Can J Diabetes 39 (2015) 26e31
Table 1 Comparative factor analysis
Factor 1 Autonomic and neuroglycopenic symptoms
Factor 2 Behavioural symptoms
Factor 3 General malaise symptoms
Table 2 Sample characteristics Retrospective recall
Prospective report
Variable
Sweating Warmth Confusion Trembling Restlessness Slurred speech Weakness Pounding heart Yawning Tingling lips Daydreaming Blurred vision Argumentativeness, defiance Irritability Aggressiveness Uncoordination Odd behaviour Feeling awful Dizziness Sleepiness Hunger Tummy pain Tiredness Feeling sick Headache Tearfulness
Sweating Warmth Confusion Trembling Restlessness Slurred speech Weakness Pounding heart Uncoordination Feeling sick Odd behaviour Dizziness Argumentativeness Defiance Irritablility Aggressiveness Headache Tearfulness Sleepiness Hunger Tiredness Tummy pain Feeling awful Yawning
Children (n¼66)
Adults (n¼53)
Proportion of males (%) Age (years) at assessment (mean SD) Age (years) at diagnosis (mean SD) Hypothyroid (%) Hypertension (%) Nephropathy (%) Retinopathy (%) Neuropathy (%) Number of hypoglycemic episodes in 4 weeks >2 hypoglycemic episodes/week Lifelong history of severe hypoglycemia Ability to tell if has hypoglycemia A1C (%, normal <6.1%) Total daily dose (units/kg, mean SD) Insulin type (%) Rapid-acting analogue Long-acting analogue Continued insulin infusion pump
49.2% males 12.12.4
41.2% males 38.714.5
NS p<0.001
7.52.9
17.512.9
p<0.001
1.6% 0% 0% 0% 0% 8.06.5
20% 14% 8% 34% 20% 10.511.0
p<0.001 p<0.01 p<0.05 p<0.001 p<0.001 NS
35.4% 59.4%
41.5% 66.7%
NS NS
92% 8.41.2 0.990.37
86% 7.81.3 0.660.26
NS p<0.05 p<0.001
74.2% 0% 12.1%
62.3% 15.1% 9.4%
p<0.001
episodes in each individual, this analysis was done using the first prospectively reported episode only. A comparative factor analysis was used to determine whether the 28 symptoms co-segregated into the 3 groups observed by Deary et al (7) in both adults and children. The equality of the factor structures across groups (i.e. those obtained for adults and those obtained for children) was determined using multisample confirmatory factor analysis with factor loadings >.3 as the cut-off. The same strategy was used to compare retrospective and prospective symptom reports. When conducting comparative factor analysis from both retrospective recall and prospective report of symptoms, 3 similar but nonidentical factors emerged, as seen in Table 1. Results Participants included 66 children (49.2% males; mean age ¼ 12.1 2.4 years; mean age at diagnosis ¼ 7.52.9 years) through parent report and 53 adults (41.2% males; mean age ¼ 38.714.5 years; mean age at diagnosis ¼ 17.512.9 years) with type 1 diabetes. Of the children, 53 (80%) were younger than 13 years of age, and 13 (20%) were 13 years of age or older. Additional characteristics of the sample are shown in Table 2. The groups differed in the duration of disease and the presence of complications. Children and adults did not differ in their abilities to know when their blood sugar levels were low or in their A1C levels, frequency of hypoglycemia or history of severe hypoglycemia. No participants reported having severe hypoglycemia during the month of observation. Children, however, had significantly higher scores for being tearful (p<0.01), being aggressive (p<0.05), having abdominal pain (p<0.01), being hungry (p<0.01) and being dizzy (p<0.05) than did adults. Adults had significantly higher scores for sweating (p<0.01), daydreaming (p<0.001), and having tingling lips (p<0.01) as compared to children. Analysis comparing childrens’ and adults’ retrospective recall of symptoms As shown in Figure 1, reported intensity of symptoms was compared between the retrospective reports and the prospective
NS, Not significant.
reports (the first episodes called in). Results showed that for adults, significantly higher scores emerged in retrospective data for all symptoms except 3: tearfulness, headache and abdominal pain. A similar pattern emerged when children were analyzed separately: significantly higher scores for retrospective data for all 28 symptoms except blurred vision and tummy pain. When conducting comparative factor analysis from the retrospective recall of symptoms, adults and children differed significantly on all 3 factors: adults reported higher scores on factor 1 (F[1,173] ¼ 15.67, p<0.001) whereas children had higher scores on factor 2 (F[1,173] ¼ 4.58, p<0.05) and factor 3 (F[1,173] ¼ 35.92, p<0.001). When the factor analysis was repeated after separating children into 2 age groups (<13 years and 13 years), only 2 distinct factors emerged (37.8% of variance explained) for the younger group, whereas 3 factors emerged (41.5% of variance explained) in the older group (data not shown). Analysis comparing childrens’ and parents’ retrospective recall of symptoms For each symptom, the rating given by the child was correlated with the rating given by the parents. Parents’ reports were significantly correlated with childrens’ reports for every symptom, with the exception of sleepiness, nightmares, slurred speech, blurry vision, tummy pain, warmth, restlessness, daydreaming, pounding heart, tingling lips and feeling awful (data not shown). Analyses using first prospectively reported episode The most common symptoms for the first episode called in for the study of adults were hunger, sweating, trembling and weakness. For children, weakness, trembling and hunger were more common. Figure 2 shows the comparison between children and adults of the symptom scores on the first prospective episode reported. The 2 most discriminating symptoms for adults vs. children for the first episode called in for the study were sleepiness and tiredness (overall model for discriminant function analysis, c2 (28, N ¼ 119) ¼ 58.27, p<0.01). Overall, 81.1% of adults and 82.5% of
A. Amin et al. / Can J Diabetes 39 (2015) 26e31
29
A
B
Figure 1. Retrospective and Prospective Symptom Reporting. Panel (A), adults; panel (B), children. The intensity of symptoms was significantly different between the retrospective recall and the prospective report, except for those marked with n/s (nonsignificant; p>0.05).
children were correctly classified as adults (or children). Being tired and feeling sleepiness were significantly more intense and more common in children (p<0.01). In a comparative factor analysis using all 28 symptoms for the first episode called in with oblique rotation, 3 factors emerged that explained 42.7% of the variance (see Table 1). As observed in the retrospective comparison, the prospective reports showed that hypoglycemic symptoms in children and adults differed significantly in terms of behavioural symptoms (factor 2) and general malaise symptoms (factor 3) but not in terms of autonomic and neuroglycopenic symptoms (factor 1). Behavioural and general malaise symptoms were significantly more common and more intense in children than in adults (F[1, 113]¼ 6.72, p<0.05 and F [1, 113] ¼ 4.64, p<0.05, respectively). The correlations between blood glucose levels and each of the 3 factor scores were not significant for adults or for children. Thus, in these 2 groups, the severity of symptoms was not linked to the actual blood glucose levels. Analyses using number of hypoglycemic episodes When the sample was stratified based on the number of hypoglycemic episodes experienced per week (i.e. <1/week, 1 to 3/week, >3/week), the largest proportion of adults (39%) and the majority of children (52%) had 1 to 3 episodes per week. The association between the number of episodes and whether the subject was a child or an adult was not significant. As shown in Figure 3, as expected, there was an inverse relationship between A1C levels and the number of hypoglycemic episodes per week in both
children and adults. The correlation between A1C levels and number of hypoglycemic episodes per week was significant for adults (r¼.378, p¼0.023), and the same pattern, albeit a trend, emerged for children (r¼.264, p¼0.084). The correlation did not differ significantly between adults and children (p¼0.589). For adults, there was a significant correlation between more hypoglycemic episodes per week and lower ratings of hunger (r¼.422, p¼0.003) but higher ratings for feelings of sickness (r¼.297, p¼0.040) and tearfulness (r¼.318, p¼0.028). For children, there was a significant correlation between more hypoglycemic episodes per week and better ability to tell whether having hypoglycemia (r¼.314, p¼0.014). Discussion and Conclusions We present the findings of this study believing it to be unique in comparing directly the differences in hypoglycemia symptomatology between children and adults with type 1 diabetes. Furthermore, only a few studies (15,17) have compared patients’ recall of hypoglycemia with actual events. This study also directly compared patients’ reported symptoms both retrospectively (i.e. recall) and prospectively (in-the-moment reporting) in contrast to past studies that examined the accuracy of reporting of severe hypoglycemic events rather than actual symptoms. Both children and adults had higher symptom scores by recollection than were achieved when they reported events during the study. This finding suggests that when hypoglycemic, patients with type 1 diabetes may have fewer symptoms than they think they have, with the potential that they are inclined to delay treatment.
30
A. Amin et al. / Can J Diabetes 39 (2015) 26e31
Figure 2. Comparison of symptoms scores on the first prospective event between adults and children. n/s: nonsignificant difference between children and adults.
It may be, however, that enrolment in the study, with detailed enquiry about their symptoms and knowing they had to report each episode for a month, they became sensitized and identified episodes earlier. Even if this latter explanation holds true, it supports the benefit of hypoglycemia awareness training to minimize the risk for severe hypoglycemic episodes (18,19). As such, providers should strive to work with patients with type 1 diabetes and their caregivers to identify the less intense symptoms that can serve as warnings for hypoglycemic episodes and enable patients and caregivers to treat hypoglycemic episodes early and reduce the risk for severe hypoglycemia. The results of this study also verify that presenting symptoms of hypoglycemia in children differ from those in adults. Tupola et al (20), using diaries of symptoms of hypoglycemia over a 3-month period in 161 children with type 1 diabetes, found that during mild episodes, the predominant symptoms were nonspecific in 41% of episodes, neuroglycopenic in 20% of episodes and autonomic in 39%
Figure 3. A1C for children and adults by number of hypoglycemic episodes per week. Note that only 44 of 66 children and 36 of 53 adults had A1C levels that were available within 3 months of the study.
of episodes. In children younger than 6 years, only 34% of episodes had predominantly autonomic symptoms compared with 57% of episodes in adolescents 15 years of age and older (p¼0.01). The most common symptoms reported in that study were weakness, tremor, hunger and drowsiness. Similarly, in our study, children more commonly reported weakness, trembling and hunger as symptoms of hypoglycemia. The symptoms of being tired and sleepy were also found to be significantly more intense and common in children than in the adults in our study. McCrimmon et al (11) interviewed 100 parents and 43 children with diabetes. They found that both autonomic and neuroglycopenic symptoms seem to be present at the same time in children. This was in contrast to adults, in whom autonomic and neuroglycopenic symptoms were well identified as being separated or hierarchical events, depending on the severity of the hypoglycemia. Our analysis failed to replicate this finding in adults; however, that may be due to the method used for glucose measurement (home glucose meter measurements vs. laboratory measurements; see discussion below in the Limitations section). McCrimmon et al (11) also reported that both parents and children identified behavioural changes as frequent and early signs of hypoglycemia. The results of our study corroborate the findings of McCrimmon et al, in that behavioural and general malaise symptoms were significantly more common and intense in children than in adults. One could postulate that the reported increase in behavioural symptoms in children is the result of an observer bias (symptoms of children are reported by parents; adults self-report). However, when we compared the symptoms reported by the children with those reported by their parents, we found a good correlation between children’s and parents’ reports for most symptoms, including behavioural symptoms. This difference between adults and adolescents may be explained in part by the differing responses to hypoglycemia seen in children compared to adults. Jones et al (10) studied the counterregulatory hormone and symptomatic responses to hypoglycemia in 13 poorly controlled children with type 1 diabetes and compared the responses to those of 16 children and 19 adults without diabetes. In children with diabetes, hypoglycemic symptoms were
A. Amin et al. / Can J Diabetes 39 (2015) 26e31
identified at glucose levels of 4.2 mmol/L compared to 3.6 mmol/L in children without diabetes and less than 3.0 mmol/L in adults without diabetes. While experiencing low blood glucose levels, the 2 groups of children had significantly higher symptom scores than did the adult group. The epinephrine and growth hormone responses paralleled the symptom scores, with more rapid and greater responses in children with diabetes. Both groups of children had significantly higher responses compared to the adult group at the end of the study. Ross et al also found that although symptom responses were similar in prepubertal, midpubertal and postpubertal children with type 1 diabetes, prepubertal children had lower blood glucose levels when autonomic reaction was observed (21). They also noted that the sweating and pancreatic polypeptide responses to hypoglycemia present in adults were seen only in postpubertal children. Limitations of this study include the nature of children’s maturation, which may not allow prepubertal children to comprehend and express some symptoms of hypoglycemia. The real-life setting of this study also limited our ability to have accurate laboratory blood glucose levels measured, which may explain the lack of hierarchy or progression of symptoms with decreasing blood glucose levels that may be observed in laboratory settings. Home glucose monitors are known be less accurate in lower glucose ranges than in normal or hyperglycemic ranges. Another limitation of this study is the difference between groups in duration of diabetes and presence of microvascular complications. However, if this difference had an effect, it might reduce the frequency or severity of autonomic symptoms in the adult group and would bias against finding a difference between the children and the adults. Finally, the number of missing A1C results limits the generalization of our results associated with this parameter. We also recognize that the nature of hypoglycemic symptoms can vary or fluctuate in individual patients with diabetes (perhaps related to altered neurohormonal responses after preceding episodes of hypoglycemia). This is, in part, the reason for our design of examining the patient recall (retrospective) of symptoms as well as the current state (prospective). Overall, this study corroborates findings that hypoglycemia symptomatology differs significantly between adults and children and notes the importance of improved recognition and treatment of hypoglycemia in both of these groups. Healthcare providers should be aware of the differing presentations of hypoglycemia between adults and children and should consider these differences when educating patients and their caregivers in developing individualized care plans. Providers should also encourage patients and their caregivers to perform regular blood glucose testing and monitoring to prevent, detect and enable early treatment of hypoglycemia. Additionally, providers should work with children with type 1 diabetes to ensure that they are aware of how their hypoglycemia symptoms will evolve as they transition between childhood and adulthood so they can move away from dependency on their caregivers to recognize hypoglycemia symptoms and reach self-recognition of hypoglycemia symptoms.
31
Acknowledgements We are grateful to the families and patients who agreed to participate in the study. This study was supported by a grant from the Stewart Fund for Diabetes Research. All authors had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. References 1. Davis E, Keating B, Byrne G, Russell M, Jones T. Impact of improved glycaemic control on rates of hypoglycaemia in insulin-dependent diabetes mellitus. Arch Dis Child 1998;78:111e5. 2. Clayton D, Woo V, Yale J. Hypoglycemia. Can J Diabetes 2013;37(Suppl 1): S69e71. 3. Diabetes Control and Complications Trial Research Group. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med 1993; 329:977e86. 4. Cryer PE. The barrier of hypoglycemia in diabetes. Diabetes 2008;57: 3169e76. 5. Hepburn DA, Deary IJ, Frier BM, et al. Symptoms of acute insulin-induced hypoglycemia in humans with and without IDDM factor-analysis approach. Diabetes Care 1991;14:949e57. 6. Hepburn DA, Dary IJ, Frier BM. Classification of symptoms of hypoglycemia in insulin-treated diabetic patients using factor analysis: Relationship to hypoglycemia unawareness. Diabet Med 1992;9:70e5. 7. Deary IJ, Hepburn DA, MacLeod KM, Brier BM. Partitioning the symptoms of hypoglycemia using multi-sample confirmatory factor analysis. Diabetologia 1993;36:771e7. 8. Diabetes Control and Complications Trial Research Group. Hypoglycemia in the diabetes control and complications trial. Diabetes 1997;46:271e86. 9. Diabetes Control and Complications Trial Research Group. Effect of intensive diabetes treatment on the development and progression of long-term complications in adolescents with insulin-dependent diabetes mellitus: Diabetes Control and Complications Trial. J Pediatr 1994;125:177e88. 10. Jones T, Boulware S, Kraemer D, et al. Independent effects of youth and poor diabetes control on responses to hypoglycemia in children. Diabetes 1991;40: 358e63. 11. McCrimmon R, Gold A, Deary I. Symptoms of hypoglycemia in children with IDDM. Diabetes Care 1995;18:858e61. 12. Graveling AJ, Noyes KJ, Allerhand MH, et al. Prevalence of impaired awareness of hypoglycemia and identification of predictive symptoms in children and adolescents with type 1 diabetes. Pediatr Diabetes 2014;15:206e13. 13. Pacaud D. Hypoglycemia: the Achilles heel of the treatment of children with type 1 diabetes. Can J Diabetes 2002;26:215e22. 14. Wherrett D, Huot C, Pacaud D. Type 1 diabetes in children and adolescents. Can J Diabetes 2013;37(Suppl 1):S153e62. 15. Gonder-Frederick LJ, Zrebiec A, Bauchowitz J, et al. Detection of hypoglycemia by children with type 1 diabetes 6 to 11 years of age and their parents: A field study. Pediatrics 2008;121:e489e95. 16. Ross LA, McCrimmon RJ, Frier BJ, et al. Hypoglycaemic symptoms reported by children with type 1 diabetes mellitus and by their parents. Diabet Med 1998; 15:836e43. 17. Davis E, Keating B, Byrne G, et al. Hypoglycemia: Incidence and clinical predictors in a large population-based sample of children and adolescents with IDDM. Diabetes Care 1997;20:22e5. 18. Cox DJ, Gonder-Frederick L, Polonsky W, et al. Blood glucose awareness training (BGAT-2): Long-term benefits. Diabetes Care 2001;24:637e42. 19. Broers S, le Cessie S, van Vliet KP, et al. Blood glucose awareness training in Dutch type 1 diabetes patients: Short-term evaluation of individual and group training. Diabet Med 2002;19:157e61. 20. Tupola S, Rajantie J. Documented symptomatic hypoglycaemia in children and adolescents using multiple daily insulin injection therapy. Diabet Med 1998; 15:492e6. 21. Ross LA, Warren RE, Kelnar CJ, Frier BM. Pubertal stage and hypoglycaemia counterregulation in type 1 diabetes. Arch Dis Child 2005;90:190e4.