Hypothesis: To address the role of continuous positive airway pressure (CPAP)on nocturnal glycemia in patients having type 1 diabetes and sleep apnea syndrome.

Investigators make the hypothesis that sleep apnea syndrome impacts nocturnal glycemia in type 1 diabetic patients and that continuous positive airway pressure treatment will permit to improve the nocturnal glycemic profile.

Study design: Adult patients with type 1 diabetes will be recruited for an extensive study of sleep habits and assessment of sleep breathing disorders. When patients will present with severe sleep apnea syndrome (apnea-hypopnea index above 30 events/hour) and insufficient glycemic control (HbA1c > 7.5%), they will be randomized in continuous positive airway pressure treatment or sham-continuous positive airway pressure treatment group for three months.

Main outcome: Nocturnal glycemic control will be assessed for 5 days before and after three months of the allocated treatment.

Quality of clinical study procedures will be checked by on-site audit.

Patient self-reported past medical history will be checked by reviewing medical records.

Paper case report forms will be used.

For the baseline extensive study of sleep habits, the following data are recorded:

• Self-reported sleep diary (7 days)
• Actimetric device (7 days)
• Munich Chronotype Questionnaire
• 24h-dietary recall
• Salivary collections every 30 minutes between 7:30 pm and 22:00 pm on one night to address the dim light melatonin onset.
• Other questionnaires: Job strain questionnaire, EPICES (precarity questionnaire) and DQOL (diabetes related quality of life)
• Nocturnal oxymetry

Blood measurements: HbA1c, lipidic profile, microalbuminuria, creatinin

After baseline assessment, patients having an abnormal nocturnal oximetry will be explored by polysomnography in order to detect a sleep apnea syndrome.

Investigators anticipate, based on a previous pilot study of our group ("Borel" Diab Med 2010), that 30% of screened patients will present with sleep apnea.

If patients present with severe sleep apnea (IAH > 30 events/hour) and HbA1c > 7.5%, they will be randomized for the intervention.

Sample size: based on a previous study by "Pallayova" et al. in patients having type 2 diabetes, investigators anticipate a difference between groups of 1.3 (SD 2.07) for the mean nocturnal glycemia. For a alpha threshold of 5% and a study power of 80%, the study should include 32 patients in each group. Taking in account a possible 25% of dropout the study will need to include 40 patients by randomized arm to demonstrate an effect.

Statistical analysis: An intention-to-treat analysis will be performed. Missing data for the continuous positive airway pressure group will be replaced by the median value at three-month of the sham-continuous positive airway pressure group and reciprocally.

In a second analysis, a per-protocol analysis will be performed limited to patients having completed their treatment in allocated group and having an adherence to continuous positive airway pressure or sham-continuous positive airway pressure treatment for more than 4-hours a night.

Normality of data will be checked by Kurtosis and Skewness tests. The significance level will be set at 5%.

Inclusion Criteria:

• Adult patients with type 1 diabetes

Exclusion Criteria:

• Pregnant women
• Patient with acute metabolic disorders during the past three months
• Shift workers
• Institutionalized patient
• Enteral and parenteral nutrition