Clinical determinants of the remission phase in children with new-onset type 1 diabetes mellitus in two years of observation

Introduction

Type 1 diabetes mellitus (T1DM) is characterised by selective and progressive autoimmune destruction of pancreatic beta cells in genetically susceptible individuals [1]. The first symptoms of T1DM usually occur several years after the destruction process has started [2, 3]. The end-stage condition is a lack of endogenous insulin production, resulting in elevated blood glucose levels. Hyperglycaemia is a factor affecting the activation of the immune system, metabolic processes, and escalating oxidative stress [4]. A few decades ago it was believed that total destruction of the beta cells, once started, is inevitable; a belief which was implied by the growing need for insulin [5]. Now we know that some  cells may remain and may still be functional, even after decades of the course of the disease [6, 7]. Effective insulin therapy reduces glucose toxicity and may affect partial recovery of surviving beta cells and increased endogenous insulin produc-tion [8, 9]. This may result in partial clinical remission (CR), also known as the honeymoon phenomenon. During the partial clinical remission period the exogenous insulin requirement decreases with proper metabolic control of diabetes and the blood glucose levels are frequently stable and within the normal range. Furthermore, residual endogenous insulin secretion in patients with T1DM is associated with improved long-term glycaemic control, reduced risk of severe hypoglycaemia [10], and reduced risk for chronic microvascular complications in patients who entered partial CR [11]. According to the International Society for Paediatrics and Adolescent Diabetes (ISPAD) Clinical Practice Consensus Guidelines 2014, the partial remission phase is defined as an insulin requirement of < 0.5 units/kg of body weight per day and glycated haemoglobin (HbA1c) < 7%. Recently, insulin dose-adjusted HbA1c, defined as HbA1c (%) + 4 × (insulin dose in units/kg/24 h), has been proposed as a more specific measure of remission [12].
It is important to discover which patients are predisposed to retaining some degree of endocrine function of the pancreas for a long period of time, and if anything can be done to preserve it. Recent studies have shown that appropriate treatment and follow-up during the honeymoon have the potential to enable the prolongation of this period for years, or even to permanently stop the destruction of the remaining ß cells; hence, the renewal of interest on the subject. Moreover, several studies have found a correlation between vari-ous factors and the occurrence of CR in T1DM. In children, young age at presentation, ketoacidosis, and low C-peptide concentrations at diagnosis connected with lower BMI reduce the likelihood of a remission phase. Additionally, the factors associated with the presence of CR taken into consideration are: HbA1c and selected antibody levels [13].
Although the beneficial effects of physical activity (PA) in the general population are numerous, many recommendations about PA for patients with T1DM are based on conclusions derived from studies on patients with type 2 diabetes (T2DM) or on healthy indi-viduals. In adult patients, higher level of PA is connected with improved cardiovascular fitness, better bone-health, and enhanced psy-chological well-being [14, 15]. No study has taken into consideration the impact of physical activity (PA) level on the occurrence of clinical remission in children with T1DM.
The aim of the present study was to describe the course of clinical remission in children with new-onset T1DM, and to com-pare the selected markers affecting the presence of partial CR, focusing on physical activity level before and during T1DM.

Material and methods

For the study, we recruited 82 patients (aged 10 ±3.71 years) hospitalised in the Department of Paediatrics, Endocrinology, and Dia-betology with Cardiology Division of the Medical University of Bialystok between 2015 and 2018 due to diagnosed T1DM, according to the ISPAD criteria [16]. The disease was diagnosed in accordance with the guidelines, by clinical manifestation, high serum glucose, and the presence of autoantibodies characteristic of type 1 diabetes mellitus (ICA, IAA, and anti-GAD) [16].
All patients were treated with insulin (multiple daily injections of insulin or continuous subcutaneous insulin infusion). Partial CR during the course of diabetes type 1 was defined as an insulin requirement < 0.5 units/kg of body weight and below 7%. For the purposes of the study, partial CR at the day of discharge from hospital was defined as low insulin demand (< 0.5 units/kg of body weight). Due to short time from diagnosis the HbA1c value could not be taken as a criterion. The frequency of partial CR was evaluated at the: 3rd, 6th, 12th, and 24th month after diagnosis. At diagnosis of T1DM the analysed parameters were: blood sugar level at admission (mg/dl), BMI-SDS, recent weight loss (kg), HbA1c (%), pH, HCO3, BE, fasting C-peptide and postprandial C-peptide concentrations (ng/ml), insulin dose (IU/kg), and presence of the following antibodies: GADA, IA2A, and ICA. Our analysis included also parameters as follows: insulin dose (IU/kg), HbA1c (%), age (years), fasting C-peptide (ng/ml), and BMI-SDS in the selected time points. We also analysed the level of physical activity before and during T1DM. Based telephone conversations with parents, the study group was di-vided into three groups as follows: no physical activity, only PE (physical education), and PE and additional sport classes.
The insulin dose was assessed as daily insulin dose per kilograms of body weight (IU/kg). HbA1c was determined by high-performance liquid chromatography (HPLC) using the Bio-Rad VARIANT™ HbA1c Program (Bio-Rad Laboratories, Inc., Hercules, CA, USA), with its values represented as percentages. Reference values of HbA1c for healthy people ranged from 4.7% to 5.7% [17]. The biological material for the study was venous blood collected from fasting patients. Next, plasma was obtained from the samples, which was used to determine the C-peptide concentration. C-peptide levels were estimated by applying the commercially available elec-trochemiluminescence method (ECLIA) (Roche Diagnostics GmbH, Penzberg, Germany). Fasting C-peptide levels in healthy individu-als ranged from 0.9 to 4.0 ng/ml, and the detection limit was assessed as 0.01 ng/ml for the assay.
Basic physical examination was performed. All children underwent physical examination, and height and weight were taken in a standard way using Harpenden stadiometer and digital scale (Seca, Germany). BMI (body mass index)-SDS (standard deviation score of BMI, according to the formula: SDS-BMI= [BMI current – BMI 50 centile]/0.5 [BMI 50 centile – BMI 3rd centile]) was calculated with standard formulas, using the results of the OLAF study of Polish children [18].

Statistical analysis

Statistical analysis was performed using Statistica 12.00 software (StatSoft, Kraków, Poland). The Kolmogorov-Smirnov test of normality was used to test the distribution of variables. For normally distributed variables, the unpaired Student t-test was used, and for non-normally distributed variables the Mann-Whitney U-test was used to compare the differences between the two groups. Relations between variables of interest were assessed by Pearson’s correlation coefficient for parametric and Spearman’s rank coefficient for non-parametric data. All data are expressed as either mean ± SD or median (interquartile range). Statistical significance was determined at the level p < 0.05. The approval of the Ethical Committee in the Medical University of Bialystok was obtained. Both parents/legal guardians and children gave their written, informed consent.

Results

The general characteristics of the study group are presented in Table 1. Most often the ISPAD criteria of remission were fulfilled within the 3rd and 6th month after diagnosis: 60% and 53% followed by 27% in the 12th month and 29% after two years of follow-up. No cases of complete CR were observed. Comparing children without physical activity with those who attended additional sport classes before diagnosis, it can be seen that they had worse metabolic parameters, such as pH, HCO3, and BE (p < 0.05) (Table 2) and higher HbA1c (p = 0.046) after one year of follow-up. Children who, during the course of T1DM, participated in additional sport classes had higher postprandial C-peptide concentrations at admission than children who only participated in PE (p = 0.049). They also had lower demand for insulin after three months (p = 0.011) and after 6, 12, and 24 months (without statistical significance) (Fig. 1). Children who were more than 10 years old at diagnosis were admitted with better biochemical parameters (pH, HCO3, BE, HbA1c, p < 0.05), had higher fasting C-peptide concentration (p = 0.0006), higher postprandial C-peptide concentration (p = 0.0004) (Fig. 2), and higher C-peptide concentration at two-year observation (p = 0.0004) than children who were less than 10 years old when diagnosed (Table 3). We obtained the same results when we compared children at age less than 10 years with children older than 15 years. Children older than 15 years had higher fasting C-peptide concentration (0.49 ±0.28 vs. 0.68 ±0.25 ng/ml, p = 0.044), postprandial C-peptide concentration (1.12 ±0.5 vs. 2.17 ±1.5 ng/ml, p = 0.001) at diagnosis, and higher fasting C-peptide concentration at two-year observation (0.25 ±0.31 vs. 0.78 ±0.17 ng/ml, p = 0.00) than children below 10 years old. There were no significant differences in C-peptide concentrations at diagnosis in the group of children aged 11-14 and ≥ 15 years. The children in whom diagnosis of T1DM was accompanied by acidosis and worse condition had lower fasting C-peptide concentration (p = 0.02), lower postprandial C-peptide concentration (p = 0.018), and also higher demand for insulin at the day of discharge (p = 0.042) and after 6 months (0.005) as well as higher HbA1c level after three months (0.026) than children who were admitted without acidosis (Table 4). Children with higher level (above median = 11.3%) at admission had significantly higher insulin requirement at the day of discharge (p = 0.003) and three months after diagnosis (p = 0.015). We observed also a negative correlation between the postprandial C-peptide level and HbA1c at diabetes diagnosis (p = 0.032).
Patients with remission at discharge from hospital presented fewer metabolic disturbances at admission (pH, HCO3, BE, p < 0.05) and lower insulin requirements from the date of discharge (p = 0.00) to the end of the first year (p < 0.05) (Table 5).
In our study we observed a positive correlation between BMI-SDS at diabetes diagnosis and fasting C-peptide concentration (p = 0.025). We compared patients with BMI-SDS higher than median BMI-SDS (–0.3) with those who had BMI-SDS lower than median BMI-SDS. Our results showed that the first group presented significantly higher fasting (p = 0.003) and postprandial C-peptide concentrations (p = 0.03) as well as lower insulin demand at the day of discharge (p = 0.01), after three months (p = 0.001), and 6 months after diagnosis (p = 0.03), and higher C-peptide in 2-year observation (p= 0.01) (Table 6).

Discussion

The prevalence of the remission phase in T1DM detected in different studies varies widely (30-80%), which partly reflects the use of different definitions (as mentioned above) [19, 20]. On average, the phase tends to appear approximately three months after the be-ginning of the insulin therapy. After that, the remission rate declines with the duration of the disease: 0-20% at 6 months and only 0-10% at 12 months [21]. In our group, at discharge from hospital 42% of patients had insulin demand < 0.5 U/kg of body weight and present-ed fewer metabolic disturbances at admission (HCO, BE) and lower insulin requirements from the date of discharge (p = 0.00) to the end of the first year. Three months after diagnosis partial remission was observed in 60% of patients, after 6 months in 53% of individu-als, after one year in 27% of them, and in 29% of them after 2 years of follow-up. The fact that there were no case of complete remission in the study group confirms that it occurs as rarely as mentioned in another study [22]. In another Polish study CR was observed in 61.83% (115/186) of patients, and often the criteria of remission were fulfilled within the 3rd and 6th month after diagnosis: 36.56% (68/186) and 39.25% (73/186), followed by 24.19%, 13.44%, and 12.37% in the 12th, 18th, and 24th months, respectively [23].
A physically active lifestyle is important in the prevention and treatment of many chronic diseases and conditions. The inde-pendent effect of physical activity in diabetes prevention is not known, but its role as a key component of lifestyle intervention has been widely described [24]. In children with T1DM, this level of physical activity can benefit glycaemic control, insulin sensitivity, protect against cardiovascular disease, and improve body composition, quality of life, and lifelong health [25]. On the other hand, due to an impaired blood glucose regulatory capability, patients with T1DM are at increased risk of adverse response to exercise when compared with healthy individuals. However, the risks of physical activity can be largely controlled with careful screening, pre-exercise prepara-tion, and appropriate exercise prescription [26]. As mentioned, patients with T1DM who are physically more active have a lower overall risk of cardiovascular events than their sedentary counterparts; therefore, this long-term benefit more than compensates for the minor increase in acute risk when physical activity programs are started [27]. Furthermore, current evidence suggests that the acute risks of physical activity-related adverse events are low [26]. However, there are still no data on the impact of physical activity level on the oc-currence of clinical remission in paediatric patients with type 1 diabetes. In our study we evaluated the level of physical activity before diagnosis and during the course of diabetes based on retrospectively collected data. Our results showed that children who participated in additional sports classes during the course of T1DM had higher postprandial C-peptide concentration than children who only took part in physical education classes at school. They also had significantly lower demand for insulin after three months as well as after 6, 12, and 24 months (but without statistical significance). When we compared children who undertook no physical activity with those who attended additional sports classes before diagnosis, the results indicated that they had worse metabolic parameters at admission to hospi-tal (pH, HCO3, BE) and higher HbA1c after one year of follow-up. The results of our study suggest that doing exercise around the time of T1DM diagnosis could have an impact on better metabolic status of patients at diabetes diagnosis and extend the “honeymoon” period of low insulin requirement.
The preservation of residual -cell function, measured by C-peptide level, is clinically important because it is associated with reduced blood glucose fluctuations and has a protective effect on both acute and long-term diabetes complications [29]. In our study children aged over 10 years at time of diagnosis had higher fasting and postprandial C-peptide levels as well as higher C-peptide con-centration after two years of observation compared to subjects younger than 10 years. Our observation was consistent with the results of Szypowska et al., who found also that children > 10 years old had higher fasting C-peptide concentration, and moreover their findings were based on a study group with 1098 Polish children [30]. The results of a study performed by Samuelsson et al. also confirmed our findings [31]. In general, prepubertal children have lower measurable C-peptide at the time of diagnosis than older individuals. While normative data in younger healthy children is not readily available, most consider this lower value to reflect that adult  cell mass is not reached until early adolescence [32]. Our, as well as the other mentioned authors’, observations may suggest a rapid and more extensive destruction of  cells or a reduced capacity to regenerate  cells in very young children.
We observed also a negative correlation between the postprandial C-peptide level and HbA1c at diabetes diagnosis. In the men-tioned study performed by Szypowska et al. a negative correlation was noted between the fasting C-peptide concentration and HbA1c at diabetes diagnosis [30]. In line with other studies, we noted statistically lower fasting (and postprandial) C-peptide levels in children with diabetic ketoacidosis compared to those without [30, 33]. When diagnosed with diabetes, if the patient presents with severe ketoac-idosis, it seems to negatively impact the onset and duration of remission (49.5% of the patients without DKA had partial remission com-pared to only 18.3% of patients with DKA) [34, 35]. This fact is in accordance with other studies from the USA and Kuwait, which also denoted this correlation [13, 34]. Increased awareness of T1DM symptoms as well as improved screening and diagnostic tools are im-portant to preserve C-peptide levels.
In our study we observed a positive correlation between BMI-SDS at diabetes diagnosis and fasting and postprandial C-peptide concentrations at the same time. We compared patients with BMI-SDS higher than median BMI-SDS (–0.3) with those who had BMI-SDS lower than median BMI-SDS. Our results showed that the first group presented significantly higher fasting and post-prandial C-peptide concentrations as well as lower insulin demand for 6 months after diagnosis. Yu et al. showed that the preserved C-peptide level increased 5-fold in overweight or obese (compared to underweight) children [36]. Redondo et al. observed that preserved C-peptide increased 2.4-fold and 4.1-fold in overweight and obese children, respectively, compared to lean children [37]. Moreover, Sosenko et al. also noted that C-peptide measures are strongly and independently related to BMI-SDS and age at, and soon after, the diagnosis of T1DM [38]. Pyziak et al. observed in their study an association between higher BMI-SDS and partial CR occurrence. The results seem to suggest that obesity/being overweight or having smaller loss of body mass in patients at T1DM diagnosis may predis-pose to further development of partial CR, and it was observed even in the youngest children below five years of age [23]. Because hyperglycaemia occurs when the amount of insulin secretion is unable to meet insulin demand, an insulin-resistant individual will pre-sent with hyperglycaemia with more residual insulin secretion than someone who is insulin sensitive. This is probably the explanation for those with higher body mass index presenting with higher C-peptide levels at the time of diagnosis [32]. However, the impact of BMI on the rate of C-peptide concentration decline during the course of diabetes type 1 remains unknown.

Conclusions

Clinical remission after onset of T1DM in children is connected with older age, lower HbA1c level, higher body mass, and better clinical condition at the time of disease recognition. Higher PA affects better biochemical parameters at diagnosis of T1DM, and lower insulin demand with higher C-peptide level during the course of diabetes. Defining the factors determining the occurrence of CR is very important, not only for the present but also for future paediatric patients with T1DM. The presence of the honeymoon phenomenon can, however, give benefits, including the search for novel therapeutic options to prolong the period of CR in children with newly diagnosed T1DM.

References

1. Odegaard JI, Chawla A. Connecting type 1 and type 2 diabetes through innate immunity. Cold Spring Harb Perspect Med 2012; 2: a007724. doi: 10.1101/cshperspect.a007724
2. Moole H, Moole V, Mamidipalli A, et al. Spontaneous complete remission of type 1 diabetes mellitus in an adult – review and case report. Community Hosp Intern Med Perspect 2015; 5: 28709. doi: 10.3402/jchimp.v5.28709
3. Abdul-Rasoul M, Habib H, Al-Khouly M. ‘The honeymoon phase’ in children with type 1 diabetes mellitus: frequency, dura-tion, and influential factors. Pediatric Diabetes 2006; 7: 101-107. doi: 10.1111/j.1399-543X.2006.00155.x
4. Brownlee M. The pathobiology of diabetic complications. Diabetes 2005; 54: 1615-1625.
5. Faustman D. Why were we wrong for so long? The pancreas type 1 diabetic patients commonly function for decades. Diabeto-logia 2014; 57: 1-3. doi: 10.1007/s00125-013-3104-9
6. Keenan HA, Sun JK, Levine J, et al. Residual insulin production and pancreatic -cell turnover after 50 years of diabetes: Joslin Medalist Study. Diabetes 2010; 59: 2846-2853. doi: 10.2337/db10-0676
7. Oram RA, McDonald TJ, Shields BM, et al. Most people with long-duration type 1 diabetes in a large population based study are insulin microsecretors. Diabetes Care 2015; 38: 323-328. doi: 10.2337/dc14-0871
8. Oram RA, Jones AG, Besser RE, et al. The majority of patients with long-duration type 1 diabetes are insulin microsecretors and have functioning beta cells. Diabetologia 2014; 57: 187-191. doi: 10.1007/s00125-013-3067-x
9. Sorensen JS, Johannesen J, Pociot F, et al. Residual beta-cell function 3 ±6 years after onset of type 1 diabetes reduces risk of severe hypoglycemia in children and adolescents. Diabetes Care 2013; 36: 3454-3459. doi: 10.2337/dc13-0418
10. Sherry NA, Tsai EB, Herold KC. Natural history of beta-cell function in type 1 diabetes. Diabetes 2005; 54: 32-39.
11. Niedzwiecki P, Pilacinski S, Uruska A, et al. Influence of remission and its duration on development of early microvascular complications in young adults with type 1 diabetes. J Diabetes Complications 2015; 29: 1105-1111. doi: 10.1016/j.jdiacomp.2015.09.002
12. Kordonouri O, Klingensmith G, Knip M, et al. ISPAD Clinical Practice Consensus Guidelines 2014. Other complications and diabetes-associated conditions in children and adolescents. Pediatr Diabetes 2014; 15: 270-278. doi: 10.1111/pedi.12183
13. Sokolowska M, Chobot A, Jarosz-Chobot P. The honeymoon phase – what we know today about the factors that can modulate the remission period in type 1 diabetes. Pediatr Endocrinol Diabetes Metab 2016; 22: 66-70. doi: 10.1111/pedi.12183
14. Chimen M, Kennedy A, Nirantharakumar K, et al. What are the health benefits of physical activity in type 1 diabetes mellitus? A literature review. Diabetologia 2012; 55: 542-551. doi: 10.1007/s00125-011-2403-2
15. Brazeau AS, Leroux C, Mircescu H, et al. Physical activity level and body composition among adults with type 1 diabetes. Dia-bet Med 2012; 29: 402-408. doi: 10.1111/j.1464-5491.2012.03757.x
16. Craig ME, Jefferies C, Dabelea D, et al. Definition, epidemiology and classification of diabetes in children and adolescents. Pe-diatric Diabetes 2014; 15 (suppl. 20): 4-7. doi: 10.1111/j.1399-5448.2009.00568.x
17. Baranowska-Jazwiecka A, Mianowska B, Fendler W, et al. Fetal hemoglobin and hemoglobin A1c level among pediatric patients with type 1 diabetes. Pediatr Endocrinol Diabetes Metab 2016; 22: 48-53. doi: 10.18544/PEDM-22.02.0050
18. Kulaga Z, Rozdzynska A, Palczewska I, et al. Percentile charts of height, body mass, and body mass index in children and ado-lescents in Poland – results of OLAF study. Standardy Medyczne/Pediatria 2010; 7: 690-700.
19. Muhammad BJ, Swift PGF, Raymond NT, et al. Partial remission phase of diabetes in children younger than age 10 years. Arch Dis Childhood 1999; 80: 367-369. doi: 10.1136/adc.80.4.367
20. Abdul-Rasoul M, Habib H, Al-Khouly M. “The honeymoon phase” in children with type 1 diabetes mellitus: frequency, dura-tion, and influential factors. Pediatr Diabetes 2006; 7: 101-107. doi: 10.1111/j.1399-543X.2006.00155.x
21. Moole H, Moole V, Mamidipalli A, et al. Spontaneous complete re- mission of type 1 diabetes mellitus in an adult – review and case report. Community Hosp Intern Med Perspect 2015; 5: 28709. doi: 10.3402/jchimp.v5.28709
22. Bonfanti R, Bognetti E, Meschi F, et al. Residual beta-cell function and spontaneous clinical remission in type 1 diabetes melli-tus: the role of puberty. Acta Diabetol 1998; 35: 91-95.
23. Pyziak A, Zmyslowska A, Bobeff K, et al. Markers influencing the presence of partial clinical remission in patients with newly diagnosed type 1 diabetes. J Pediatr Endocrinol Metab 2017; 30: 1147-1153. doi: 10.1515/jpem-2017-0100
24. Colberg SR, Laan R, Dassau E, et al. Physical activity and Type 1 diabetes. J Diabetes Sci Technol 2015; 9: 609-618. doi: 10.1177/ 1932296814566231
25. Quirk H, Glazebrook C, Blake H. A physical activity intervention for children with type 1 diabetes – steps to active kids with diabetes (STAK-D): a feasibility study. BMC Pediatrics 2018; 18: 37. doi: 10.1186/s12887-018-1036-8.
26. Burr JF, Shephard RJ, Riddell MC. Physical activity in type 1 diabetes mellitus: assessing risks for physical activity clearance and prescription. Can Fam Physician 2012; 58: 533-535.
27. Herbst A, Kordonouri O, Schwab KO, et al. Impact of physical activity on cardiovascular risk factors in children with type 1 diabetes. Diabetes Care. 2007; 30: 2098-2100. doi: 10.2337/dc06-2636
28. Ludvigsson J, Carlsson A, Deli A, et al. Decline of C-peptide during the first year after diagnosis of type 1 diabetes in children and adolescents. Diabetes Res Clin Pract 2013; 100: 203-209. doi: 10.1016/j.diabres.2013.03.003
29. Kuhtreiber WM, Washer SL, Hsu E, et al. Low levels of C-peptide have clinical significance for established type 1 diabetes. Di-abet Med 2015; 32: 1346-1353. doi: 10.1111/dme.12850
30. Szypowska A, Groele L, Wysocka-Mincewicz M, et al. Factors associated with preservation of C-peptide levels at the diagnosis of type 1 diabetes. J Diabetes Complications 2018; 32: 570-574. doi: 10.1016/j.jdiacomp.2018.03.009
31. Samuelsson U, Lindblad B, Carlsson A, et al. Residual beta cell function at diagnosis of type 1 diabetes in children and adoles-cents varies with gender and season. Diabetes Metab Res Rev 2013; 29: 85-89. doi: 10.1002/dmrr.2365
32. VanBuecken DE,Greenbaum CJ. Residual C-peptide in type 1 diabetes: what do we really know? Pediatr Diabetes 2014; 15: 84-90. doi: 10.1111/pedi.12135
33. Szypowska A, Dżygało K, Wysocka-Mincewicz M, et al. High incidence of diabetic ketoacidosis at diagnosis of type 1 diabetes among Polish children aged 10–12 and under 5 years of age: a multicenter study. Pediatr Diabetes 2017; 18: 722-728. doi: 10.1111/pedi.12446
34. Bowden SA, Duck MM, Hoffman RP. Young children (< 5 yr) and adolescents (> 12yr) with type 1 diabetes mellitus have low rate of partial remission: diabetic ketoacidosis is an important risk factor. Pediatric Diabetes 2008; 9: 197-201. doi: 10.1111/j.1399-5448.2008.00376.x
35. Neylon OM, White M, O Connell MA, et al. Insulin-dose adjusted HbA1c-defined partial remission phase in a pediatric popu-lation – when is the honeymoon over? Diabetic Medicine 2013; 30: 627-628. doi: 10.1111/dme.12097
36. Yu HW, Lee YJ, Cho WI, et al. Preserved C-peptide levels in overweight or obese compared with underweight children upon di-agnosis of type 1 diabetes mellitus. Ann Pediatr Endocrinol Metab 2015; 20: 92-97. doi: 10.6065/apem.2015.20.2.92
37. Redondo MJ, Rodriguez LM, Escalante M, et al. Beta cell function and BMI in ethnically diverse children with newly diagnosed autoimmune type 1 diabetes. Pediatr Diabetes 2012; 13: 564-571. doi: 10.1111/j.1399-5448.2012.00875.x
38. Sosenko JM, Geyer S, Skyler JS. The influence of body mass index and age on C-peptide at the diagnosis of type 1 diabetes in children who participated in the diabetes prevention trial-type 1. Pediatr Diabetes 2018, 19: 403-409. doi: 10.1111/pedi.12609