Non-Inferiority of Insulin Glargine Versus Insulin Detemir on Blood Glucose Variability in Type 1 Diabetes Patients: A Multicenter,Randomized,Crossover Study

Abstract

Background:

This study compared the effects of insulin glargine and insulin detemir on blood glucose variability under clinical practice conditions in patients with type 1 diabetes (T1D) using glulisine as the mealtime insulin.

Methods:

This was a multicenter, crossover trial in 88 randomized T1D patients: 54 men and 34 women, 46.8±13.7 years old, with a duration of diabetes of 18±9 years and hemoglobin A1c level of 7.1±0.7%. The per-protocol population included 78 patients: 44 received glargine/detemir and 34 detemir/glargine in the first/second 16-week period, respectively. The primary end point was the coefficient of variation (CV) of fasting blood glucose (FBG). Secondary end points included variability of pre-dinner blood glucose, mean amplitude of glycemic excursions, mean of daily differences, and doses and number of daily insulin injections. The non-inferiority criterion was an insulin glargine/insulin detemir FBG CV ratio with a 95% confidence interval (CI) upper limit ≤1.25.

Results:

The non-inferiority criterion was satisfied with a mean value of 1.016 (95% CI=0.970–1.065). Intention-to-treat analysis confirmed the non-inferiority with a 95% CI upper limit=1.062. No significant differences were found on secondary objectives, but there was a trend to higher doses and number of daily injections with insulin detemir. A total of eight (four glargine and four detemir) patients reported nine serious adverse events (including one severe episode of hypoglycemia). None of them was considered as related to basal insulins. Serious adverse events led to treatment discontinuation in two patients of the detemir group and none in the glargine group.

Conclusions:

In T1D patients under clinical practice conditions, insulin glargine was non-inferior to insulin detemir regarding blood glucose variability, as assessed by CV of FBG.

Introduction

I n patients with type 1 diabetes (T1D), similar doses of subcutaneous insulin injections may lead to considerable intra- and inter-individual differences in blood glucose control.^1
–3 This hampers practical insulin therapy and may also play an independent role in the development of diabetes complications,^4

–11 although this point is still a subject of controversy.¹²

The long-acting insulin analog glargine induces less blood glucose variability than NPH insulin in adult healthy volunteers¹³ and in adult and pediatric subjects with T1D.^13

–16 Similarly, the long-acting insulin analog detemir has been associated with less glycemic variability than NPH insulin in T1D patients.^17
–19 The only reported study comparing the intra-individual variability of insulin glargine and insulin detemir found a significantly more reproducible effect on blood glucose with insulin detemir;²⁰ however, the authors compared the effects of both insulins under experimental conditions, by measuring the intra-individual variability of the glucose infusion rates required to maintain euglycemic glucose clamp conditions (target blood glucose concentration 5.5 mmol/L) in T1D patients. Insulin detemir showed a significantly lower glucose infusion rate coefficient of variation (CV) (27%) than insulin glargine (46%).²⁰

The above arguments prompted us to compare insulin detemir and insulin glargine by measuring their direct effects on fasting blood glucose variability in T1D patients, under conditions of clinical practice. Insulin glulisine was added as mealtime insulin to both treatment groups (basal–bolus regimen).

Patients and Methods

This was a multicenter, randomized, crossover study designed to investigate the within-subject glycemic variability in subjects with T1D using insulin glargine or insulin detemir as basal insulin. The study protocol was reviewed and approved by the local ethics committee and was performed according to the French rules of good clinical practice (Huriet Law, modified by French Law umber 2004-806) and the Declaration of Helsinki.

After giving signed informed consent, 135 adult T1D patients from 25 diabetes care centers were included in the trial. T1D was defined by a C-peptide concentration of <0.1 nmol/L and a fasting blood glucose (FBG) ≥7 mmol/L. Inclusion criteria were a duration of diabetes of ≥3 years, hemoglobin A1c (HbA1c) ≤8.5% at the inclusion visit, and intensive insulin therapy treatment for at least 6 months by using a basal–bolus regimen with insulin glargine as basal (evening) insulin. Carbohydrate counting for the adjustment of meal insulin doses was not routinely performed by most of the included patients, whereas dose adjustment according to capillary blood glucose measurements was a common practice.

Patients continued their current insulin treatment for 1–2 weeks and then received glulisine as prandial insulin (three times per day) for an initial period of 4 weeks. Then, patients with a more than 50% of pre-dinner blood glucose (PDBG) level of ≤8.3 mmol/L during the last 3 weeks of the initial period were randomized in two crossover groups using insulin glargine or insulin detemir. Each crossover period lasted 16 weeks, without washout between both periods.

The change of insulin treatment was done as follows: (1) switch of fast-acting insulin to insulin glulisine, dose per dose; (2) switch of insulin glargine to insulin detemir, dose per dose at the same hour in the evening; and (3) switch of insulin detemir to insulin glargine (a) for patients receiving one daily insulin detemir injection, dose per dose at the same evening hour, or (b) for patients receiving two daily insulin detemir injections, insulin glargine was started in the evening, at 80% of the total daily dose of detemir.

The insulin doses were titrated for each patient based on the following blood glucose targets: (1) fasting and before meals, 5.0 mmol/L<blood glucose ≤7.2 mmol/L; (2) 1–2 h after meal starting, blood glucose <9.9 mmol/L; and (3) at bedtime (at least 2.5 h after the last meal), 6.1 mmol/L<blood glucose ≤8.3 mmol/L.

The adjustment of insulin doses was performed according to the following protocols. Evening basal insulin was adapted every 3–4 days according to fasting blood glucose levels: ≤90 mg/dL or hypoglycemia due to basal insulin, −2 IU; 91–130 mg/dL, no change; 131–150 mg/dL, +2 IU; 151–180 mg/dL, +3 IU; and >180 mg/dL, +4 IU. Premeal insulin was adapted every 1–2 days according to 1- or 2-h postmeal blood glucose levels: ≤130 mg/dL, −1 IU; 131–179 mg/dL, no change; 180–210 mg/dL, +1 IU; 211–240 mg/dL, +2 IU; and >240 mg/dL, +3 IU.

Insulin glargine was always given as one daily (evening) injection. Conversely, patients who failed to reach the PDBG target with a single daily (evening) injection of insulin detemir were switched to two daily injections (by adding one pre-breakfast injection).

The study included a total of 11 consultations and five telephone contacts between the investigator and the patient. Patients collected daily FBG and PDBG, if possible during the entire two treatment periods, and for at least the last 8 weeks of each period.

Evaluation criteria

The primary end point was the CV of FBG during the last 8 weeks of each crossover period.

Secondary end points were CV of PDBG measured during the last 8 weeks of each crossover period; daily blood glucose profiles recorded at seven points (before breakfast, 2 h after breakfast, before lunch, 2 h after lunch, before dinner, 2 h after dinner, and at bedtime) during 2 consecutive days, before the last two visits of each crossover period; mean amplitude of glycemic excursions (MAGE), the mean value of glycemic excursions higher than 1 SD,² recorded over 2 days before the last two visits of each crossover period; mean of daily differences (MODD), the mean value of differences in blood glucose between two consecutive days (at the same time of the day),² recorded before the last two visits of each crossover period; HbA1c levels at the end of each treatment period; body weight at the end of each treatment period; and doses and number of daily insulin injections at the end of each treatment period.

All adverse events, including hypoglycemia, were to be reported throughout the trial. Severe hypoglycemia was defined as an episode in which the patient's condition requires the indispensable assistance of a third person and is associated with blood glucose of <1.98 mmol/L or a quick recovery after ingestion of sugar or intravenous glucose or glucagon administration. The observation period of adverse events started with the informed consent and lasted until 24 h after the final day of the trial.

Statistical analysis

Sample size calculation was based on the following assumptions: α=0.025, power (1−β)=0.95, non-inferiority limit=1.25 (ln: 0.223), true difference=1.05 (ln: 0.049), SD=0.2. With a rate of not exploitable subjects of about 15% over the two crossover periods in per-protocol (PP) analysis, the total number of subjects to be randomized was 86. Data are given as mean±SD values.

Analysis of primary end point

FBG variability was estimated by the CV (BG SD/mean BG) calculated on the data collected during the last 2 months of each crossover period (weeks 8–16 and weeks 24–32). The assumption of non-inferiority of insulin glargine compared with insulin detemir was tested by an analysis of variance model in crossover in which fixed effects were the treatment group, period, and sequence. The upper limit of the unilateral 95% confidence interval (CI) to 95% in unilateral position was calculated, and non-inferiority was established if the upper limit of the confidence interval does not exceed 125%. The aim was to demonstrate that CV_{insulin glargine} is ≤1.25×CV_{insulin detemir}. The primary analysis was conducted in the PP population, and its robustness was tested in the intention-to-treat (ITT) population.

Secondary end points

PDBG variability was analyzed by similar methods as that of FBG. The analysis of the other criteria was descriptive, without performing tests.

Results

One hundred thirty-five T1D patients were recruited (Fig. 1). Nineteen participants were excluded during the initial period because they did not replace their fast insulin with insulin glulisine. Of the remaining 116 patients who completed the initial period, 28 were not randomized because of failure to comply with the randomization criteria. Of the 88 randomized patients, 50 received insulin glargine (first period)/insulin detemir (second period), and 38 received insulin detemir (first period)/insulin glargine (second period) as basal insulin. The randomization was skewed because of the fact that it was organized per investigation center. As a consequence, it happened that in the centers that randomized few patients the allocation to glargine (first period)/detemir (second period) or detemir (first period)/glargine (second period) was not balanced. The difference between this trial distribution (50:38) and a balanced one (44:44) was not statistically significant (by Fisher's test, P value=0.241). Ten patients were further excluded from the analysis because of major protocol deviations: crossover period duration <3 months (n=8) or number of FBG measurements <42 during at least one a crossover period (n=2). The PP population thus included 78 T1D patients: 44 patients received insulin glargine (first period)/insulin detemir (second period), and 34 patients received insulin detemir (first period)/insulin glargine (second period) as basal insulin.

FIG. 1.

Patient recruitment and follow-up during the study. D, detemir; G, Glargine; ITT, intention-to-treat; PP, per-protocol.

There were no clinically relevant differences in demographic characteristics between subjects in the two treatment groups (Table 1). PP patients were predominantly males, with long mean duration of diabetes (17–18 years) and mean Hb1Ac values slightly higher than 7%. They had received insulin glargine as basal insulin for about 3 years.

Table 1.

Baseline Characteristics of the Two Treatment Groups According to the Order of Basal Insulin Use (First/Second) During the Study (Per-Protocol Population)

	Usage order
	Insulin glargine/insulin detemir	Insulin detemir/insulin glargine
Number of patients	44	34
Age (years)	48.3±13.6	46.4±14.1
Males (%)	65.9	55.9
BMI (kg/m²)	24.6±3.5	25.3±3.5
Duration of diabetes (years)	17.1±8.4	18.5±10.1
HbA1c (%)	7.06±0.69	7.16±0.71
Fasting blood glucose (mg/dL)	140±68	130±59
Pre-dinner blood glucose (mg/dL)	141±67	124±48
Blood pressure (mm Hg)
Systolic	127.3±12.1	123.1±11.4
Diastolic	74.7±7.0	72.0±8.0
Complications of diabetes (% of patients)
Retinopathy	20.5	26.5
Neuropathy	22.7	11.8
Duration of insulin glargine treatment (months)	35.3±16.8	32.2±10.4

Data are mean±SD values or percentage of patients as indicated.

BMI, body mass index; HbA1c, hemoglobin A1c.

Non-inferiority of insulin glargine on glycemic variability

Primary criteria

The CV of FBG was not significantly different between insulin glargine (41.1±12.0%) and insulin detemir (39.9±10.9%) (P=0.494) (Table 2). Maximum and minimum FBG CV values were also similar for both treatments (i.e., maximum FBG CV 66.9% and 66.8% for insulin glargine and insulin detemir, respectively, and minimum FBG CV 7.3% and 9.3% for insulin glargine and insulin detemir, respectively). Differences remained non-significant when corrected for period (P=0.1116) or sequence of treatment (P=0.962). Mean value of the insulin glargine/insulin detemir FBG CV ratio was 1.016 (95% CI 0.970–1.065). The upper limit of the CI 95% for the insulin glargine/insulin detemir FBG CV ratio (1.065) was lower than 1.25, satisfying the non-inferiority criterion. This was confirmed by (1) the ITT analysis, showing an upper limit of the 95% CI for the insulin glargine/insulin detemir FBG CV ratio of 1.062, a value still lower than 1.25, and (2) the observation that the upper limit of the 95% CI for the ln (FBG CV_glargine – CV_detemir) (0.063) was <0.223 (ln 1.25).

Table 2.

Glycemic Variability

	Insulin glargine (n=44)	Insulin detemir (n=34)	P
FBG CV (%)	41.1±12.0	39.9±10.9	0.49
PDBG CV (%)	40.1±11.2	38.9±11.3	0.29
MAGE (g/L)	1.01±0.67	0.97±0.47	ND
MODD (g/L)	0.55±0.27	0.55±0.27	ND

Data are mean±SD values.

CV, coefficient of variation; FBG, fasting blood glucose; MAGE, mean amplitude of glucose excursions; MODD, mean of daily differences; ND, not determined; PDBG, pre-dinner blood glucose.

Secondary criteria

The non-inferiority of insulin glargine on glycemic variability was further confirmed by other glycemic data. CV of PDBG levels was not significantly different between insulin glargine (40.1±11.7%) and insulin detemir (38.9±11.3%) (P=0.494). The differences remained nonsignificant even when corrected by period (P=0.695) or sequence of treatment (P=0.619), and the insulin glargine/insulin detemir PDBG CV ratio (1.025; 95% CI 0.979–1.073) was lower than 1.25 (upper limit for non-inferiority).

The MAGE and the MODD were also similar in both treatment groups.

Other secondary end points

HbA1c

Changes in HbA1c levels were small (≤0.20%) and similar between both treatment groups. HbA1c decreased by 0.19±0.34% during the first insulin glargine period and by 0.10±0.52% during the second insulin glargine period, whereas it decreased by 0.20±0.55% during the first insulin detemir period and by 0.14±0.38% during the second insulin detemir period.

Body weight

Changes in body weight were small and similar between both treatment groups. Mean value of body weight was unchanged during insulin glargine treatment and decreased by 0.2 kg during insulin detemir treatment (difference not significant).

Insulin doses

At the end of each treatment period, the daily dose of insulin glargine was lower than that of insulin detemir (median, 0.28 U/kg and 0.39 U/kg, respectively). The daily doses of insulin glulisine were similar at the end of each treatment period, with their median values ranging from 0.32 U/kg to 0.35 U/kg regardless of the basal insulin.

Number of insulin injections

During the last month of each crossover period, 60% of patients in the insulin detemir group received two daily insulin detemir injections to achieve the PDBG target, whereas only once-daily injection was permitted for insulin glargine.

Safety

Both insulin treatments were generally well tolerated. The frequency of adverse events was similar for insulin glargine and insulin detemir (32.9% and 36.0%, respectively). They were essentially co-morbid conditions, associated with or promoted by diabetes, particularly infections (17.6% and 23.3% of patients during insulin glargine and insulin detemir periods, respectively). Only three adverse events were related (according to investigators' opinion) to basal insulin: weight gain (1.2%) on insulin glargine in one patient and signs of local intolerance on insulin detemir in two patients.

A total of eight (four glargine and four detemir) patients reported nine serious adverse events. None of them was considered related to insulin glargine or insulin detemir. Serious adverse events led to treatment discontinuation in two patients of the detemir group and none in the glargine group. One severe hypoglycemia episode was reported as a serious adverse event in the insulin glargine group but was assessed as related to insulin glulisine.

No deaths occurred during the study.

Hypoglycemia

The median monthly rate of symptomatic hypoglycemia was broadly similar in the insulin glargine group (2.32) and in the insulin detemir group (2.16). The frequency of severe symptomatic hypoglycemia tended to be higher with insulin glargine (10 patients [11.8%]) compared with insulin detemir (four patients [4.7%]), but the difference was not statistically significant (by Fisher's Exact Test, P=0.1023). Most episodes occurred during the titration period (weeks 1–8).

Discussion

This crossover trial in 88 T1D patients demonstrated the non-inferiority of insulin glargine on FBG variability compared with insulin detemir when used in clinical practice conditions. Furthermore, insulin glargine showed no inferiority concerning other measurements of glucose variability, such as the CV of PDBG and the MAGE and the MODD indices. Our results contrast with those of Heise et al.,²⁰ who found a much lower reproducibility of insulin glargine compared with insulin detemir in a comparable population of T1D patients. One possible explanatory factor may be that Heise et al.²⁰ assessed the within-subject variability under euglycemic glucose clamp conditions on four study days, a condition far from “real life.” The mealtime doses of insulin that were administered in our trial could also interact with the effects of the basal insulin on blood glucose variability, so that in common practice the observed blood glucose variability results from the overall insulin therapy. Moreover, the effects of any basal insulin can be only assessable in common practice after the installation of a certain steady state of plasma insulin levels, which is unlikely achieved in 4 days. Pieber et al.²¹ have reported a head-to-head 26-week comparison of glargine and detemir administered randomly in two parallel groups of T1D patients, in combination with insulin aspart as fast-acting analog. Although this trial focused on HbA1c and hypoglycemia occurrence, results on within-subject variability of premeal blood glucose values were reported. No significant difference on pre-breakfast and pre-lunch plasma glucose variations was revealed, whereas pre-dinner variations were marginally lower in the detemir group (P=0.049). On the other hand, our results are in line with those of Heller et al.²² showing that a 52-week treatment with insulin glargine or insulin detemir (in combination with insulin aspart as meal insulin) did not differ in HbA1c levels, tolerability, and occurrence of hypoglycemia.

Some limitations of our trial have to be acknowledged in terms of external validity. First, we included only patients treated with insulin glargine as basal insulin, thereby excluding patients who might have failed in obtaining acceptable blood glucose control with insulin glargine and switched to insulin detemir. Second, we did not randomize 28 among the 116 included patients because more than 50% of their PDBG values exceeded 8.3 mmol/L during the run-in phase. These non-randomized patients could have benefited from a twice-daily injection of insulin detemir.

Of note is that in our study glucose variability was high in patients of both basal insulin treatment groups (about 40% for FBG and PDBG, vs. about 31% and 34% in the study of Pieber et al.²¹), to which the short-acting insulin analog may have contributed. Furthermore, there were wide inter-individual variations in CV (ranging from 6.3% to 66.9%), perhaps related to the patient population heterogeneity in duration of diabetes, total daily insulin dosage, and/or body mass index. These observations point out the progress still needed to be achieved both in the development of new insulin formulations and in patient education for diabetes management.

Previous studies in T1D patients suggested that persistent wide fluctuations in plasma glucose, in the presence of low mean glucose and HbA1c values, were involved in the increased risk of severe hypoglycemia with intensive insulin therapy and the necessity to adapt insulin doses.^14,23 Here, the similar glycemic variability was associated with a median rate of hypoglycemia that was roughly similar with insulin glargine (2.32 events/month) and insulin detemir (2.16 events/month). Severe hypoglycemia occurred in 10 patients with insulin glargine and in four patients with insulin detemir, but the difference was not statistically significant. Most occurred during the titration phase.

In conclusion, insulin glargine is not inferior to insulin detemir in a basal–bolus insulin treatment of T1D patients concerning blood glucose variability, as assessed by CV of FBG. Both basal insulins are well tolerated, with a similar broad rate of hypoglycemia and a comparable metabolic effect as measured by HbA1c and weight variations.

Appendix

The members of the Variability Study Group include: Eric Renard and Florence Galtier (Montpellier University Hospital, Montpellier, France), Didier Gouet, Thierry Godeau, and Marc Bernardin (Saint Louis Hospital, La Rochelle, France), Marc Lévy (Max Fourestier Hospital, Nanterre, France), Pierre Sérusclat (Minguettes Polyclinic, Vénissieux, France), Dominique Paris-Bockel, Nathalie Jeandidier, and Sylvie Boullu-Sanchis (Civil Hospital, Strasbourg, France), Brigitte Kuchly-Anton (Nantes University Hospital, Nantes, France), Catherine Petit (Gilles de Corbeil Hospital, Corbeil-Essonnes, France), Sylvie Pêcheur, Blandine Gatta, and Cédric Fagour (Haut-Lévêque Hospital, Bordeaux, France), Bruno Guerci (Hospital Jeanne d'Arc, Centre Hospitalier Universitaire of Nancy, Toul, France), Danièle Dubois-Laforgue (Cochin Hospital, Paris, France), Annie Clergeot-Begey (Dijon University Hospital, Dijon, France), Yannick Lorcy and Anne-Marie Leguerrier (Rennes University Hospital, Rennes, France), Jean-Pierre Courrèges (Narbonne Hospital, Narbonne, France), Loïc Geffray (Robert Bisson Hospital, Lisieux, France), Nathalie Jourdan (Nîmes University Hospital, Nîmes, France), Sylvaine Clavel (Hôtel Dieu, Le Creusot, France), Annie Fonteille (Annecy, France), Christine Rouby (Toulouse, France), Christian Alessis (Marseille, France), Ignacio Lopez (Mantes la Jolie, France), Jean-Paul Beressi (André Mignot Hospital, Le Chesnay, France), Jean Michel Darocha (Nevers Hospital, Nevers, France), Pascal Monguillon (Pasteur Clinic, Brest, France), and Sylvie Brionnet-Metzger (Clermont-Ferrand, France).

Footnotes

Acknowledgments

This study was sponsored by Sanofi-Aventis, France. The authors wish to thank the sponsor's office of statistics and Ricardo Garay for their assistance in the analyses of study data and the preparation of the manuscript, respectively.

Author Disclosure Statement

Honoraria were paid by Sanofi-Aventis to the investigating teams for their participation in the study.

References

Heinemann

. Variability of insulin absorption and insulin action. Diabetes Technol Ther, 2002; 4:673–682.

Guerci

, Sauvanet

. Subcutaneous insulin: pharmacokinetic variability and glycemic variability. Diabetes Metab, 2005; 31:4S7–4S24.

Gin

, Hanaire-Broutin

. Reproducibility and variability in the action of injected insulin. Diabetes Metab, 2005; 31:7–13.

Monnier

, Mas

, Ginet

, Michel

, Villon

, Cristol

J-P

, Colette

. Activation of oxidative stress by acute glucose fluctuations compared with sustained chronic hyperglycemia in patients with type 2 diabetes. JAMA, 2006; 295:1681–1687.

Brownlee

. The pathobiology of diabetic complications: a unifying mechanism. Diabetes Care, 2005; 54:1615–1625.

Temelkova-Kurktschiev

, Koehler

, Henkel

, Leonhardt

, Fuecker

, Hanefeld

. Postchallenge plasma glucose and glycemic spikes are more strongly associated with atherosclerosis than fasting glucose or HbA1c level. Diabetes Care, 2000; 23:1830–1834.

Esposito

, Giugliano

, Nappo

, Marfella

. Regression of carotid atherosclerosis by control of postprandial hyperglycemia in type 2 diabetes mellitus. Circulation, 2004; 110:214–219.

Ceriello

, Cavarape

, Martinelli

, Da Ros

, Marra

, Quagliaro

, Piconi

, Assaloni

, Motz

. The post-prandial state in type 2 diabetes and endothelial dysfunction: effects of insulin aspart. Diabet Med, 2004; 21:171–175.

Weber

, Schnell

. The assessment of glycemic variability and its impact on diabetes-related complications: an overview. Diabetes Technol Ther, 2009; 11:623–633.

10.

Brownlee

, Hirsch

. Glycemic variability: a hemoglobin A1c-independent risk factor for diabetic complications. JAMA, 2006; 295:1707–1708.

11.

Kilpatrick

, Rigby

, Atkin

. A1C variability and the risk of microvascular complications in type 1 diabetes: data from the Diabetes Control and Complications Trial. Diabetes Care, 2008; 31:2198–2202.

12.

Kilpatrick

, Rigby

, Atkin

. The effect of glucose variability on the risk of microvascular complications in Type 1 diabetes. Diabetes Care, 2006; 29:1486–1490.

13.

Gerich

, Becker

RHA

, Zhu

, Bolli

. Fluctuation of serum basal insulin levels following single and multiple dosing of insulin glargine. Diabetes Technol Ther, 2006; 8:237–243.

14.

White

, Chase

, Arslanian

, Tamborlane

. for the 4030 Study Group. Comparison of glycemic variability associated with insulin glargine and intermediate-acting insulin when used as the basal component of multiple daily injections for adolescents with type 1 diabetes. Diabetes Care, 2009; 32:387–393.

15.

Bolli

, Songini

, Trovati

, Del Prato

, Ghirlanda

, Cordera

, Trevisan

, Riccardi

, Noacco

. Lower fasting blood glucose, glucose variability and nocturnal hypoglycaemia with glargine vs NPH basal insulin in subjects with Type 1 diabetes. Nutr Metab Cardiovasc Dis, 2009; 19:571–579.

16.

Raskin

, Klaff

, Bergenstal

, Halle

, Donley

, Mecca

. A 16-week comparison of the novel insulin analog insulin glargine (HOE 901) and NPH human insulin used with insulin lispro in patients with type 1 diabetes. Diabetes Care, 2000; 23:1666–1671.

17.

Vague

, Selam

, Skeie

, De Leeuw

, Elte

JWF

, Haahr

, Kristensen

, Draeger

. Insulin detemir is associated with more predictable glycemic control and reduced risk of hypoglycemia than NPH insulin in patients with type 1 diabetes on a basal-bolus regimen with premeal insulin aspart. Diabetes Care, 2003; 25:590–596.

18.

Hermansen

, Madsbad

, Perrild

, Kristensen

, Axelsen

. Comparison of the soluble basal insulin analog insulin detemir with NPH insulin. Diabetes Care, 2001; 24:296–301.

19.

Hermansen

, Fontaine

, Kukolja

, Peterkova

, Leth

. Insulin analogues (insulin detemir and insulin aspart) versus traditional human insulins (NPH insulin and regular human insulin) in basal-bolus therapy for patients with type 1 diabetes. Diabetologia, 2004; 47:622–625.

20.

Heise

, Nosek

, Rønn

, Endahl

, Heinemann

, Kapitza

, Draeger

. Lower within-subject variability of insulin detemir in comparison to NPH insulin and insulin glargine in people with type 1 diabetes. Diabetes, 2004; 53:1614–1620.

21.

Pieber

, Treichel

, Hompesch

, Philotheou

, Mordhorst

, Gall

, Robertson

. Comparison of insulin detemir and insulin glargine in subjects with Type 1 diabetes using intensive insulin therapy. Diabet Med, 2007; 24:635–642.

22.

Heller

, Koenen

, Bode

. Comparison of insulin detemir and insulin glargine in a basal–bolus regimen, with insulin aspart as the mealtime insulin, in patients with Type 1 diabetes: a 52-week, multinational, randomized, open-label, parallel-group, treat-to-target noninferiority trial. Clin Ther, 2009; 31:2086–2097.

23.

Reach

. Continuous glucose monitoring and diabetes health outcomes: a critical appraisal. Diabetes Technol Ther, 2008; 10:69–80.