Sleep for Health Study on the Effects of Cognitive Behavioral Therapy for Insomnia on Diabetes Risk

Type 2 diabetes mellitus (T2D) is a major cause of blindness, kidney failure, cardiovascular disease, amputations, reduced quality of life, and premature death in the United States, and it is expected that one in three Americans will have T2D by 2050. To stem the tide of this health crisis, new strategies are needed to prevent the progression to T2D from prediabetes-elevated glucose levels that are not yet in the diabetes range. A growing body of research suggests that insomnia is a major modifiable risk factor for progression to diabetes. The proposed study would build off a promising feasibility study to test whether providing cognitive behavioral therapy for insomnia (CBT-I) to patients with prediabetes results in a reduction in glucose levels compared to a patient education control program. If so, this insomnia treatment could be an effective tool to prevent diabetes.

Individuals with prediabetes and insomnia will be randomized to receive six sessions of a deployment-ready digital CBT-I program, providing standard-of-care treatment for insomnia (intervention arm, n = 150), or a patient education website providing nontailored material about insomnia (control arm, n = 150). The investigators will complete assessments at baseline, at 11 weeks (after the conclusion of the intervention and control programs), and at 33 weeks post-baseline, measuring hyperglycemia, objective and subjective measures of sleep, and potential mediating variables including diet, exercise, and mood.

The investigators will assess (1) whether individuals randomized to the intervention arm have lower rates of hyperglycemia, as measured by oral glucose tolerance testing and various secondary measures, than individuals randomized to the control arm at 11 weeks and 33 weeks after baseline; (2) whether improvements in sleep after baseline are associated with decreases in hyperglycemia, regardless of study arm; and (3) whether any effects of the intervention on hyperglycemia are mediated by improvements in sleep, diet, exercise, and/or mood.

This research will serve as a critical step in identifying a potentially dramatic tool for improving health outcomes for Americans at risk of T2DM. Sleep interventions can lead to sustained improvements that are intrinsically rewarding to patients. If effective, digital CBT-I could provide a powerful pathway to preventing diabetes for millions of patients with prediabetes.

Inclusion criteria

• Age ≥ 22 years and < 80 years of age
• Prediabetes
• Insomnia
• Regular access to device with internet access
• Adequate data at baseline

Exclusion criteria

• BMI > 40 kg/m2
• Sleep comorbidities detected in medical record or via medical history
• Shift work or significant, externally imposed irregular sleep schedule
• moderate to severe OSA by home sleep apnea test as part of trial protocol
• Received a full course of CBT-I in the last 12 months
• Current use of medication with glycemic effects:
• History of type 1 or type 2 diabetes or recent/planned use of hypoglycemic agents (e.g., metformin, insulin)
• Recent history of bariatric surgery or planning bariatric surgery in the next year
• Current or recent use of weight loss meds
• Unstable sleep medication regimen (recent change to schedule or dosage)
• Significant comorbidity that may interfere with CBT-I uptake or increase risks
• Unwilling or unable to limit heavy machinery use/long bouts of driving or unstable illness that would be worsened by sleep restriction
• High risk of falls
• Epilepsy
• Medical conditions that interfere with dCBT-I or contribute to insomnia or diabetes risk (e.g., hyperthyroidism, significant kidney disease, active cancer treatment, any medical condition that requires chronic steroid use)
• Significant alcohol or substance use disorder
• Active or recent history of eating disorder, recent weight change of >10%
• Women: pregnancy (current or planned), breastfeeding, < 1 year postpartum
• Use of hydroxyurea
• Extensive skin changes or adhesive allergy making CGM sensor use problematic