A randomized trial comparing the rate of hypoglycemia-assessed using continuous glucose monitoring-in 125 preschool children with type 1 diabetes treated with insulin glargine or NPH insulin (the PRESCHOOL study)

Pediatric DiabetesVolume 14, Issue 8 p. 593-601 Original ArticleOpen Access A randomized trial comparing the rate of hypoglycemia—assessed using continuous glucose monitoring—in 125 preschool children with type 1 diabetes treated with insulin glargine or NPH insulin (the PRESCHOOL study) Correction(s) for this article Corrigendum Volume 16Issue 6Pediatric Diabetes pages: 462-462 First Published online: August 11, 2015 Thomas Danne, Corresponding Author Thomas Danne Kinder- und Jugendkrankenhaus "AUF DER BULT", Hannover, Germany Corresponding author: Prof. Thomas Danne, MD, Kinder- und Jugendkrankenhaus "AUF DER BULT", Diabetes Centre for Children and Adolescents, Janusz-Korczak-Allee 12, 30173 Hannover, Germany. Tel: +49 511 8115 3330; fax: +49 511 8115 5334; e-mail: danne@hka.deSearch for more papers by this authorAreti Philotheou, Areti Philotheou Diabetes Clinical Trials Unit, UCT Private Academic Hospital, Cape Town, South AfricaSearch for more papers by this authorDavid Goldman, David Goldman Sanofi, Bridgewater, NJ, USASearch for more papers by this authorXiang Guo, Xiang Guo Sanofi, Bridgewater, NJ, USASearch for more papers by this authorLin Ping, Lin Ping Sanofi, Bridgewater, NJ, USASearch for more papers by this authorAnna Cali, Anna Cali Sanofi, Paris, FranceSearch for more papers by this authorPeter Johnston, Peter Johnston Sanofi, Bridgewater, NJ, USASearch for more papers by this author Thomas Danne, Corresponding Author Thomas Danne Kinder- und Jugendkrankenhaus "AUF DER BULT", Hannover, Germany Corresponding author: Prof. Thomas Danne, MD, Kinder- und Jugendkrankenhaus "AUF DER BULT", Diabetes Centre for Children and Adolescents, Janusz-Korczak-Allee 12, 30173 Hannover, Germany. Tel: +49 511 8115 3330; fax: +49 511 8115 5334; e-mail: danne@hka.deSearch for more papers by this authorAreti Philotheou, Areti Philotheou Diabetes Clinical Trials Unit, UCT Private Academic Hospital, Cape Town, South AfricaSearch for more papers by this authorDavid Goldman, David Goldman Sanofi, Bridgewater, NJ, USASearch for more papers by this authorXiang Guo, Xiang Guo Sanofi, Bridgewater, NJ, USASearch for more papers by this authorLin Ping, Lin Ping Sanofi, Bridgewater, NJ, USASearch for more papers by this authorAnna Cali, Anna Cali Sanofi, Paris, FranceSearch for more papers by this authorPeter Johnston, Peter Johnston Sanofi, Bridgewater, NJ, USASearch for more papers by this author First published: 03 June 2013 https://doi.org/10.1111/pedi.12051Citations: 22 The copyright line for this article was changed on 25 March 2015 after original online publication AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat Abstract Background Avoidance of hypoglycemia is a key consideration in treating young children with type 1 diabetes (T1DM). Key Objective To evaluate hypoglycemia with insulin glargine vs. neutral protamine Hagedorn (NPH) insulin in young children, using continuous glucose monitoring (CGM). Subjects Children of 1 to <6 yr treated with once-daily glargine vs. once- or twice-daily NPH, with bolus insulin lispro/regular human insulin provided to all. Methods Twenty-four week, multicenter, randomized, open-label study. Primary endpoint was event rate of composite hypoglycemia [symptomatic hypoglycemia, low CGM excursions (<3.9 mmol/L) or low fingerstick blood glucose (FSBG; <3.9 mmol/L)]. Noninferiority of glargine vs. NPH was assessed for the primary endpoint. Results One hundred and twenty-five patients (mean age, 4.2 yr) were randomized to treatment (glargine, n = 61; NPH, n = 64). At baseline, mean HbA1c was 8.0 and 8.2% with glargine and NPH, respectively. Composite hypoglycemia episodes/100 patient-yr was 1.93 for glargine and 1.69 for NPH; glargine noninferiority was not met. Events/100 patient-yr of symptomatic hypoglycemia were 0.26 for glargine vs. 0.33 for NPH; low CGM excursions 0.75 vs. 0.72; and low FSBG 1.93 vs.1.68. There was a slight difference in between-group severe/nocturnal/severe nocturnal hypoglycemia and glycemic control. All glargine-treated patients received once-daily injections; on most study days NPH-treated patients received twice-daily injections. Conclusions While glargine noninferiority was not achieved, in young children with T1DM, there was a slight difference in hypoglycemia outcomes and glycemic control between glargine and NPH. Once-daily glargine may therefore be a feasible alternative basal insulin in young populations, in whom administering injections can be problematic. Increases in the prevalence of type 1 diabetes mellitus (T1DM) among those <18 yr of age have been reported in Europe 1, 2, the USA 3, 4, and elsewhere 5. In particular, a doubling of newly diagnosed cases in children <5 yr of age is expected by 2020 in Europe 1. In young children ≤6 yr of age, the management of T1DM is a challenge, as the rate of treatment-related hypoglycemia in this cohort is more than twofold higher than that observed among older children 6, 7. This may be partly due to increased physical activity and irregular dietary patterns that have been identified as risk factors for hypoglycemia in children and adolescents with T1DM 6. While the health risk of long-term hyperglycemia is much greater than any lasting consequences of hypoglycemia, the fear of hypoglycemia is stressful for parents of children with T1DM 8. Young children neither tolerate hypoglycemia well—seizures 6, 9, cognitive impairment 10, and structural neurologic abnormalities are not uncommon 11—nor do they verbalize the symptoms well. Moreover, almost 60% of young children experience impaired awareness of hypoglycemia, which is associated with a 1.6-fold increased risk of severe hypoglycemia 9. Consequently, avoidance of hypoglycemia—particularly severe hypoglycemia—is a key consideration in the treatment of young children with T1DM. Continuous glucose monitoring (CGM) may be useful in detecting episodes of potential or actual hypoglycemia. The PRESCHOOL study used CGM to assess the occurrence of hypoglycemia in children <6 yr of age with T1DM receiving insulin lispro or regular human insulin at mealtimes, plus one of two basal insulins: once-daily prebreakfast insulin glargine or once- or twice-daily neutral protamine Hagedorn (NPH) insulin. Insulin glargine is a once-daily basal insulin with a flat plasma pharmacokinetic profile that has shown greater blood glucose stability, lower fasting glucose, and a lower risk of hypoglycemia, especially nocturnal hypoglycemia, in adult patients with T1DM compared with basal insulins with pharmacokinetic peaks 12. Retrospective data have also shown that insulin glargine reduced the risk of severe hypoglycemia (especially nocturnal) compared with NPH insulin in 128 children with T1DM aged <6 yr 13. Prospective data of this kind in large groups of young children with T1DM, however, are lacking. Therefore, as adult data are not directly transferable to pediatric patients 14, PRESCHOOL aimed to address this knowledge gap. Methods Subjects and study design PRESCHOOL was a 24-wk, multicenter (n = 61), multinational (n = 16), randomized, open-label, parallel-group study. Children aged ≥1 to <6 yr with T1DM (≥1 yr duration) who were receiving multiple daily insulin injections for ≥2 months were included if they had glycated hemoglobin (HbA1c) values at screening of 6–12% (inclusive), and provided a family member was able to generate 6 days' worth of CGM data during the 2-wk run-in period. Subjects using anti-hyperglycemia agents other than insulin were excluded. Patients were centrally randomized in a 1:1 ratio to one of the two treatment groups. Randomization was stratified based on the number of hypoglycemic events (<0.5 vs. ≥0.5 episodes/24 h) and baseline HbA1c (<8.5% vs. ≥8.5%) during screening; stratification was expected to contribute to clearer conclusions regarding the respective efficacy variables and overall efficacy of insulin glargine. Patients were randomized to treatment with either morning insulin glargine or once- or wice-daily NPH; all received bolus insulin lispro or regular human insulin (at mealtime and/orbedtime). Insulin was titrated to achieve fasting blood glucose of 5–8.0 mmol/L, bedtime blood glucose of 6.7–10.0 mmol/L, and nocturnal blood glucose of 4.4–9.0 mmol/L. Best efforts were made to complete the up-titration of both basal insulins by week 12. Doses of insulin glargine and NPH insulin were increased no more often than once a week, but doses could be reduced due to hypoglycemia at any time. The 24-wk treatment period was preceded by a 2-wk run-in period and followed by a 2-wk post-treatment period. Clinic visits occurred at screening (week −3/–2), randomization (week 0), weeks 1, 2, 4, 6, 8, 10, 12, 16, 20, 24 (end of treatment), and 26 (follow-up). CGM data were captured at weeks 0, 2, 4, 8, 12, 16, 20, and 24 using the DexCom Seven Plus CGM system (DexCom, San Diego, CA, USA). Patients were required to use CGM for ≥6 total days for ≥6 occasions during the on-treatment period, of which two occasions had to occur after week 14. Over each CGM occasion, at least 1296 usable CGM glucose values were to be recorded. Fingerstick blood glucose (FSBG) data were captured using Roche ACCU-CHEK glucometers (Roche, Indianapolis, IN, USA) at screening and all subsequent clinic visits, while HbA1c was measured at screening and wk 12 and 24 by central laboratories blinded to treatment. Adverse events were monitored from baseline to 7 days after the last treatment visit. Ethical approval according to local regulations was obtained from independent ethics committees and/or institutional review boards for all study sites. Conduct of the study was compliant with standards of data collection for clinical trials (Clinicaltrials.gov: NCT00993473), according to the Declaration of Helsinki. Written informed consent was obtained from the parent/legal guardian of each participant. Objectives The primary efficacy endpoint was to compare the effect of once-daily insulin glargine dosed in the morning vs. once- or twice-daily NPH insulin on the composite hypoglycemia rate consisting of (i) symptomatic hypoglycemia episodes, which were recorded in patient diaries, then validated by study investigators; (ii) low CGM glucose excursions (<3.9 mmol/L), which were confirmed by FSBG <3.9 mmol/L 10 min before to 10 min after the low CGM excursion (i.e., confirmed low CGM); and (iii) FSBG <3.9 mmol/L, which was recorded ≥1 h from the end of a confirmed low CGM excursion. Double-counting of hypoglycemic events was avoided by not counting any event that occurred in the hour following an earlier, counted event. Secondary endpoints included rates of severe symptomatic hypoglycemia, nocturnal hypoglycemia, nocturnal symptomatic hypoglycemia, and severe nocturnal symptomatic hypoglycemia; HbA1c values at study end; change in HbA1c values from baseline; and HbA1c <7.5% rates at study end. Severe hypoglycemia was defined as an event requiring assistance from another person, as a result of altered consciousness, to administer carbohydrate, glucagon, or to take other actions. Nocturnal events were those beginning between 23:00 hours and 7:00 hours. Main secondary endpoints based on CGM values were (i) average daily blood glucose at study end and change from baseline; (ii) average daily blood glucose over the entire study; and (iii) blood glucose variability (including nocturnal). Statistical analyses Analysis was performed on the modified intent-to-treat population, defined as all randomized patients who received at least one dose of study medication. No per-protocol population was defined. Sample size calculation was based on an expected composite hypoglycemia rate of 0.8 events/100 patient-yr of exposure to insulin glargine or to NPH insulin. The sample size and novel composite outcome was planned to ensure sufficient power so that the upper bound of the two-sided 95% confidence interval (CI) for the insulin glargine:NPH ratio of the mean composite hypoglycemia rates for the comparison of treatment groups would not exceed 1.15. A sample size of 35 completed patients per treatment group was to provide 96% power to demonstrate noninferiority of insulin glargine vs. NPH. A sequential, stepwise, closed-testing approach was used to assess noninferiority and superiority [if the upper bound of the 95% CI for the ratio of rates of the primary composite endpoint for insulin glargine to NPH insulin was <1.15 (noninferior) or <1.0 (superior)]. One-sided tests were performed at the 0.025 level of significance. Details regarding calculation of the primary composite endpoint, and statistical models used for comparisons of the two treatment arms and endpoint analyses are described in the Appendix S1, Supporting Information. Results A total of 165 patients were screened and 125 patients were randomized to treatment; 61 to insulin glargine and 64 to NPH insulin (Fig. 1). One patient randomized to NPH insulin was actually treated with insulin glargine, thus the safety population comprised 62 patients for insulin glargine and 63 for NPH insulin. More patients in the NPH insulin group (10/64, 15.6%) than in the insulin glargine group (4/61, 6.6%) prematurely discontinued the study. The CGM device was worn for a mean of 82.5 days in the insulin glargine group and 76.2 days in the NPH group. Of the 111 completers (57/61, 93.4% using insulin glargine; 54/64, 84.4% using NPH insulin), 99 patients (48/61, 78.7% and 51/64, 79.7%, respectively) satisfied the protocol-required CGM performance. Figure 1Open in figure viewerPowerPoint Patient disposition. CGM, continuous glucose monitoring; NPH, neutral protamine Hagedorn; AEs, adverse events. Baseline characteristics were similar between groups. The mean (SD) age of participants was 4.2 (1.0) yr, and 78.4% (98/125) of patients were over 3 yr of age (Table 1). Almost all patients (115/125, 92.0%) were using basal insulin at baseline, with 33.6% (42/125) requiring ≥2 daily basal insulin injections (Table 1). A total of 56.8% (71/125) of patients were using NPH insulin, 26.4% (33/125) insulin glargine, and 14.4% (18/125) insulin detemir at baseline. By the end of the 2-wk run-in period, the majority of patients had experienced ≥1 low CGM excursion [insulin glargine, 56/61 (91.8%); NPH insulin, 59/64 (92.2%)]. Mean (SD) HbA1c was 8.0% (1.0%) in the insulin glargine group, and 8.2% (1.4%) in the NPH insulin group (Table 3). Table 1. Baseline characteristics Insulin glargine (n = 61) NPH insulin (n = 64) Total (N = 125) Age, yr Mean (SD) 4.3 (0.9) 4.1 (1.0) 4.2 (1.0) Median (range) 5.0 (2–5) 4.0 (1–6) 4.0 (1–6) Age group, n (%) ≤3 yr 10 (16.4) 17 (26.6) 27 (21.6) >3 yr 51 (83.6) 47 (73.4) 98 (78.4) Gender, n (%) Male 32 (52.5) 30 (46.9) 62 (49.6) Female 29 (47.5) 34 (53.1) 63 (50.4) Race, n (%) White 53 (86.9) 48 (75.0) 101 (80.8) Black 2 (3.3) 2 (3.1) 4 (3.2) Asian 4 (6.6) 11 (17.2) 15 (12.0) Other 2 (3.3) 3 (4.7) 5 (4.0) Diabetes duration, yr Mean (SD) 2.1 (1.2) 2.1 (1.0) 2.1 (1.1) Median (range) 1.6 (1.0–5.3) 2.1 (1.0–4.9) 1.8 (1.0–5.3) Insulin type, n (%) Short-acting 54 (88.5) 58 (90.6) 112 (89.6) Basal 58 (95.1) 57 (89.1) 115 (92.0) Premixed 5 (8.2) 8 (12.5) 13 (10.4) Number of daily basal insulin injectionsaa Basal/bolus insulin injection at baseline defined as the last day with basal/bolus insulin dose >0 U before first administration of study drug on or after randomization. , n (%) 1 32 (52.5) 41 (64.1) 73 (58.4) 2 21 (34.4) 15 (23.4) 36 (28.8) ≥3 5 (8.2) 1 (1.6) 6 (4.8) Total daily dose of basal insulin injection (U)aa Basal/bolus insulin injection at baseline defined as the last day with basal/bolus insulin dose >0 U before first administration of study drug on or after randomization. Mean (SD) 7.3 (4.1) 7.6 (4.8) 7.5 (4.4) Median (range) 6.0 (2.0–24.0) 6.0 (1.5–24.0) 6.0 (1.5–24.0) Total daily dose of bolus insulin injection (U)aa Basal/bolus insulin injection at baseline defined as the last day with basal/bolus insulin dose >0 U before first administration of study drug on or after randomization. Mean (SD) 7.1 (3.6) 8.0 (7.2) 7.6 (5.8) Median (range) 7.8 (1.3–16.0) 7.0 (0.8–45.0) 7.0 (0.8–45.0) a Basal/bolus insulin injection at baseline defined as the last day with basal/bolus insulin dose >0 U before first administration of study drug on or after randomization. The mean daily dose for insulin glargine was relatively stable over the 24-wk treatment period (week 1, 0.35 U/kg; week 24, 0.38 U/kg), whereas it increased for NPH insulin (week 1, 0.37 U/kg; week 24, 0.46 U/kg) (Fig. 2). All patients on insulin glargine used one basal injection in the morning, whereas over the entire treatment with NPH insulin, patients on NPH received two basal injections per day on more study days than one basal injection per day. The mean daily dose of bolus insulin in the two treatment groups was similar at week 1 (insulin glargine group, 0.41 U/kg; NPH insulin group, 0.42 U/kg) and remained practically unchanged at week 24 (insulin glargine group, 0.41 U/kg; NPH insulin group, 0.40 U/kg). Figure 2Open in figure viewerPowerPoint Mean (SE) daily insulin dose over 24 wk. NPH, neutral protamine Hagedorn. Hypoglycemia The incidence rate of composite hypoglycemia was 1.93 mean events/100 patient-yr for insulin glargine, and 1.69 for NPH insulin. The upper bound of the 95% CIs exceeded the prespecified noninferiority margin of 1.15 [incidence ratio estimate, 1.18 (95% CI: 0.97–1.44); Table 2]; thus, the statistical criteria for noninferiority of insulin glargine to NPH insulin were not met. Of the individual components of the composite endpoint, mean events/100 patient-yr of symptomatic hypoglycemia were 0.26 with insulin glargine and 0.33 with NPH insulin (incidence ratio estimate, 0.76 [95% CI: 0.46–1.25]), and of confirmed low CGM excursions were 0.75 vs. 0.72, respectively [incidence ratio estimate, 1.06 (95% CI: 0.79–1.42)]. The rate of low FSBG values was 1.93 with insulin glargine and 1.68 with NPH insulin [incidence ratio estimate, 1.18 (95% CI: 0.96–1.45); Table 2]. Table 2. Incidence and rates of hypoglycemia by type and treatment assignment Efficacy endpoint Insulin glargine (n = 61) NPH insulin (n = 64) Incidence ratio estimate (SE) [95% CI] Composite hypoglycemiaaa Composite outcome of symptomatic hypoglycemia, confirmed low CGM excursions (<3.9 mmol/L), and low FSBG (<3.9 mmol/L). , n (%) 61 (100) 63 (98.4) Events/100 patient-yr, mean (SD) 1.93 (1.19) 1.69 (1.01) 1.18 (0.12) [0.97–1.44] Symptomatic hypoglycemia, n (%) 40 (65.6) 44 (68.8) Events/100 patient-yr, mean (SD) 0.26 (0.37) 0.33 (0.48) 0.76 (0.19) [0.46–1.25] Confirmed low CGM, n (%) 60 (98.4) 61 (95.3) Events/100 patient-yr, mean (SD) 0.75 (0.74) 0.72 (0.53) 1.06 (0.16) [0.79–1.42] FSBG <3.9 mmol/L, n (%) 61 (100) 63 (98.4) Events/100 patient-yr, mean (SD) 1.93 (1.22) 1.68 (1.01) 1.18 (0.12) [0.96–1.45] Severe symptomatic hypoglycemia, n (%) 4 (6.6) 2 (3.1) Events/100 patient-yr, mean (SD) 0.0014 (0.0055) 0.0007 (0.0038) 1.97 (1.67) [0.37–10.37] Nocturnal hypoglycemia, n (%) 59 (96.7) 60 (93.8) Events/100 patient-yr, mean (SD) 0.34 (0.26) 0.31 (0.25) 1.09 (0.14) [0.84–1.40] Nocturnal symptomatic hypoglycemia, n (%) 17 (27.9) 28 (43.8) Events/100 patient-yr, mean (SD) 0.024 (0.054) 0.037 (0.068) 0.65 (0.25) [0.30–1.39] Severe nocturnal symptomatic hypoglycemia, n (%) 1 (1.6) 0 Events/100 patient-yr, mean (SD) 0.0004 (0.0029) 0.0000 (0.0000) – a Composite outcome of symptomatic hypoglycemia, confirmed low CGM excursions (<3.9 mmol/L), and low FSBG (<3.9 mmol/L). Very few events of severe hypoglycemia occurred; four in the glargine group and two in the NPH group. For nocturnal hypoglycemia ('all hypoglycemia' that occurred between 23:00 hours and 7:00 hours), and for nocturnal symptomatic hypoglycemia, the glargine:NPH incidence ratios were 1.09 and 0.65, respectively (Table 2). There were multiple peak occurrences of composite hypoglycemia events around mealtimes and before bedtime in the insulin glargine group, whereas a peak of events was seen around noontime in the NPH insulin group (Fig. 3A). Peaks of lesser magnitude than those in the insulin glargine group also occurred at mealtimes in the NPH insulin group. Low FSBG readings that were independent of confirmed low CGM or symptomatic hypoglycemia events occurred in both groups near times of the three main meals (when bolus insulin was dosed and FSBG was tested for bolus dosing) and near bedtime (Fig. 3B). Figure 3Open in figure viewerPowerPoint Mean (A) composite hypoglycemia events, (B) low FSBG events not related to confirmed low CGM or symptomatic hypoglycemia events, (C) CGM glucose values over 24 h, and (D) mean (SD) 24-h glucose based on CGM glucose values over 24 wk, by treatment. NPH, neutral protamine Hagedorn. More patients in the insulin glargine group (49/61, 80.3%) than in the NPH insulin group (35/64, 54.7%) had to switch to a new basal insulin regimen from their pretrial basal insulin. In addition, more FSBG was performed overall in the insulin glargine group (mean 1056 glucose values/patient) than in the NPH insulin group (909; ratio 1.16). Glycemic control Patients treated with insulin glargine showed a slight decrease from baseline in HbA1c from baseline to week 24 [(LS) least squares mean, –0.048] while a slight increase from baseline in HbA1c (LS mean, 0.045) was observed in patients treated with NPH insulin. However, the LS mean difference between the two groups (−0.09) was relatively small. The proportion of patients with HbA1c values <7.5% at study end was comparable (Table 3). Table 3. Average daily blood glucose, glucose variability, and HbA1c by treatment group Insulin glargine (n = 61) NPH insulin (n = 64) Average daily blood glucose (mmol/L), mean (SD) Baseline 11.3 (1.9) 11.2 (2.0) Study end 11.1 (2.1) 11.7 (2.2) Change from baseline −0.2 (2.4) 0.5 (1.9) Blood glucose variability (mmol/L), mean (SD) 4.9 (0.8) 5.1 (0.7) Nocturnal blood glucose variability (mmol/L), mean (SD) 4.7 (1.0) 4.8 (0.8) HbA1c (%) Baseline, mean (SD) 8.0 (1.0) 8.2 (1.4) Study end, mean (SD) 8.1 (0.9) 8.3 (1.2) Change from baseline, LS mean (SE) −0.048 (0.11) 0.045 (0.11) LS mean difference (95% CI)aa ANCOVA model with treatment and randomization strata (number of CGM hypoglycemic excursions during the 2-wk run-in period and baseline HbA1c) as fixed effects, using baseline value as covariate. −0.093 (−0.39–0.21) Patients with HbA1c <7.5%, % (n/N) 22.0 (13/59) 22.8 (13/57) Estimated difference (95% CI)bb Estimated difference and 95% CI obtained from Cochran-Mantel-Haenszel method using randomization strata. −0.6 (−15.4–14.2) LS, least squares; CI, confidence interval. a ANCOVA model with treatment and randomization strata (number of CGM hypoglycemic excursions during the 2-wk run-in period and baseline HbA1c) as fixed effects, using baseline value as covariate. b Estimated difference and 95% CI obtained from Cochran-Mantel-Haenszel method using randomization strata. Over 24 h, mean CGM glucose was lower in the insulin glargine group than in the NPH insulin group during the day, especially in the afternoon and evening, but was higher than in the NPH insulin group overnight (Fig. 3C). At study end, the average daily blood glucose was comparable between groups, albeit numerically lower in the insulin glargine group (Fig. 3D and Table 3). There was a slight difference between treatment groups in blood glucose variability or nocturnal blood glucose variability (Table 3). Safety The proportion of patients with treatment-emergent adverse events was similar with insulin glargine (40/62, 64.5%) and NPH insulin (43/63, 68.3%). Discussion PRESCHOOL is the largest prospective study to date investigating the occurrence of hypoglycemia in children with T1DM aged ≥1 to <6 yr. It compared once-daily basal insulin glargine to conventional NPH insulin therapy usually injected twice daily. The objective of the study, to compare hypoglycemia rates in this hard-to-recruit population, could only be accomplished feasibly with a composite outcome, consisting of events of symptomatic hypoglycemia, low CGM excursions, and low FSBG measurements, which was needed to provide adequate power to detect differences in hypoglycemia between regimens. Analysis of the primary efficacy endpoint did not demonstrate noninferiority of insulin glargine to NPH insulin in the rate of composite hypoglycemia. However, in terms of the more conventional metrics included in the secondary endpoints, such as symptomatic and nocturnal symptomatic hypoglycemia, as well as severe symptomatic/nocturnal symptomatic hypoglycemia and change from baseline in HbA1c, there was a slight difference in results between treatment groups. These results partially support those of Dixon et al. 13. In that retrospective study of 128 children aged <6 yr with T1DM, glycemic control between insulin glargine and NPH insulin was also comparable, although severe hypoglycemia episodes were decreased with glargine treatment. Earlier, in 2001, a study comparing glargine and NPH in 349 older children (5–16 yr old) with T1DM also demonstrated comparable between-treatment glycemic control, as well as comparable rates of symptomatic, severe and severe nocturnal hypoglycemia 15. Thus, the results of PRESCHOOL are generally consistent with those of earlier studies, although they are particularly valuable due to the study's prospective design and use of CGM in a very young population. Indeed, CGM technology is relatively new in the management of T1DM and its use is not common practice 16, 17. Therefore, recent systematic reviews and meta-analyses have highlighted that evidence from randomized trials for the effectiveness of CGM is currently lacking, especially in children (including toddlers and preschool children) with T1DM 16, 18, 19. Compared with NPH, which was predominantly injected twice daily, a once-daily injection of glargine was equally effective in terms of glycemic control, with a lower mean daily dose. A once-daily basal insulin could be particularly advantageous in young children such as these, who are receiving multiple daily injections as part of a basal-bolus regimen. Administering insulin injections in this population can result in problems such as lipodystrophy, injection-site pain, and bruising 14. Once-daily insulin glargine can significantly reduce the number of daily injections required compared with NPH, as previously demonstrated in a 2008 study 20, and thus the potential for injection-related side effects. In addition, once-daily injections are convenient, potentially aiding good compliance with insulin therapy. In the insulin glargine group, there were a greater number of low FSBG values, which were not recorded at times of low CGM excursions or symptomatic hypoglycemia signals. The majority of low FSBG readings in both groups occurred near the three main mealtimes and near bedtime, and the monitoring was probably performed as part of parents' premeal FSBG routine. It is common clinical knowledge that families introduced to a new insulin regimen test FSBG more frequently and are less likely to change the insulin dose. As more patients in the insulin glargine group had to switch to a new basal insulin regimen from their pretrial basal insulin, more FSBG was performed overall in this group (mean measurements per patient, 1056 for glargine vs. 909 for NPH). CGM data were used to gather more information on low blood glucose. The use of CGM monitoring is a more objective means of low blood glucose detection than alternative methods, as it is independent of parental choice regarding when to check FSBG instinctively, which may be influenced by other factors such as mealtimes or recent/upcoming periods of exercise. Investigation of hypoglycemia by CGM showed a slight difference between regimens in terms of confirmed low CGM. CGM also permits an accurate assessment of daily blood glucose variability, which in this study was moderate and comparable between the treatment groups. Figure 3C clearly illustrates that the curves of mean CGM glucose over 24 h show a similar amplitude of variation for both treatment groups, with the biggest changes occurring around mealtimes. This may mean that bolus insulin, and mealtime glucose swings, played a greater role than basal insulin during the daytime in blood glucose variability, a concept supported by an earlier study in children taking bolus insulin lispro 21. CGM curves also offer interesting insights into patterns of glycemic excursions in preschool children where parental concerns appear to be a major factor. Glucose levels appeared closest to target early in the morning but rose considerably after breakfast when many preschool children attend daycare away from direct parental supervision. Consistent with parental interventions is the rise in CGM glucose at 10 pm in the glargine group,