The Science of Diabetes Self-Management and Care2023, Vol. 49(1) 11 –22© The Author(s) 2022Article reuse guidelines:sagepub.com/journals-permissionsDOI: 10.1177/26350106221136495journals.sagepub.com/home/tde
Abstract
Purpose: The purpose of this study was to evaluate the feasibility and acceptability of a technology-assisted behavioral sleep intervention (Sleep-Opt-In) and to examine the effects of Sleep-Opt-In on sleep duration and regularity, glucose indices, and patient-reported outcomes. Short sleep duration and irregular sleep schedules are associated with reduced glycemic control and greater glycemic variability.
Methods: A randomized controlled parallel-arm pilot study was employed. Adults with type 1 diabetes (n = 14) were recruited from the Midwest and randomized 3:2 to the sleep-optimization (Sleep-Opt-In) or Healthy Living attention control group. Sleep-Opt-In was an 8-week, remotely delivered intervention consisting of digital lessons, sleep tracker, and weekly coaching phone calls by a trained sleep coach. Assessments of sleep (actigraphy), glucose (A1C, continuous glucose monitoring), and patient-reported outcomes (questionnaires for daytime sleepiness, fatigue, diabetes distress, and depressive mood) were completed at baseline and at completion of the intervention.
Results: Sleep-Opt-In was feasible and acceptable. Those in Sleep-Opt-In with objectively confirmed short or irregular sleep demonstrated an improvement in sleep regularity (25 minutes), reduced glycemic variability (3.2%), and improved time in range (6.9%) compared to the Healthy Living attention control group. Patientreported outcomes improved only for the Sleep-Opt-In group. Fatigue and depressive mood improved compared to the control.
Conclusions: Sleep-Opt-In is feasible, acceptable, and promising for further evaluation as a means to improve sleep duration or regularity in the population of people with type 1 diabetes.
Type 1 diabetes (T1D) affects nearly 1.6 million people in the United States.1 The incidence of T1D is increasing by approximately 3% to 4% per year2 for reasons that are not completely understood but likely include behavioral and environmental factors.3 Despite advances in diabetes management technology, which include improved insulin formulations, improved insulin delivery methods, and better means to monitor glucose levels, only 20% of adults with T1D achieve glycemic targets4 needed to prevent and reduce diabetes complications.5
Insufficient sleep (insufficient total sleep time) and irregular sleep timing (variability in the occurrence of sleep within a 24-hour period)6 are increasingly recognized as important contributors to glycemic control and variability in T1D.6-8 Up to 40% of adults with T1D had a sleep duration less than 6 to 6.5 hours per night, either by self-report or objectively assessed actigraphy.8-10 Insufficient sleep is associated with insulin resistance and poorer glycemic control in T1D.8,11 One night of sleep restriction in adults with T1D was shown to decrease the glucose disposal rate (an index of insulin sensitivity) by 14%.11 Irregular sleep schedules are also detrimental to glycemic control and can lead to circadian disruption. The circadian system plays an important role in glucose metabolism, and experimental circadian misalignment has been shown to lead to impaired glucose tolerance.12-14 Chontong et al10 reported that irregular sleep schedules were associated with significantly higher A1C levels in persons with T1D.
In addition to glycemic control (A1C), poor sleep has been associated with greater glycemic variability. Poor subjective sleep quality was associated with greater nocturnal glycemic variability15 and higher daily mean glucose.16 Sleep disruptions17,18 and objectively measured sleep deficiencies were also associated with greater glycemic variability in persons with T1D.19
Attainment of glycemic targets largely depends on the day-to-day behaviors and self-management practices of those living with the condition.20 Poor sleep can affect self-management behavior.21,22 Reasons for this include next-day fatigue and decreased motivation and memory following less-restorative sleep.21,22 McDonough et al22 reported that the frequency of next-day blood glucose checks and total daily insulin doses were increased with longer sleep duration, suggesting that previous-night sleep may affect next-day self-care.
Persons with T1D report many challenges related to sleep due to emotional adjustment and the demands of managing a complex condition. Hypoglycemia is an ever-present danger, and fear of hypoglycemia is worse at night.23 The fear and distress associated with nocturnal hypoglycemia can lead to delayed bedtimes, awakening to check blood glucose levels, and running blood glucose levels higher during the night.24 Similarly, poor sleep quality is closely linked to diabetes distress (defined as emotional distress surrounding the management of diabetes25) among adults with T1D.7 Qualitative studies indicated that sleep was a major source of distress that influenced self-management behavior, which suggests poor sleep may be part of a cycle of emotional distress and poorer disease management and outcomes.24
Despite findings that insufficient sleep and irregular sleep are common in T1D, evidence is limited for approaches to improve them. A small number of studies have been conducted in children, primarily focusing on sleep extension or sleep hygiene. Perfect et al26 demonstrated that a 6-day sleep intervention designed to extend time in bed for school-age children was feasible and resulted in an average increase of sleep duration of 40.95 minutes. Similarly, Jaser et al27 evaluated an intervention for child/parent dyads that entailed 3 brief telephone sessions to determine the effect on sleep duration, sleep quality, diabetes distress, and other psychological variables. Sleep duration improved among the parents only, while subjective sleep quality improved in both parents and children. Tumakaka et al28 evaluated the effect of a 10-minute sleep hygiene video designed for children. In a 3-day follow-up, the children’s sleep quality was improved compared to a control group. These strategies are promising but may not be applicable to adults and have not addressed irregular sleep timing. Due to this gap, a technology-assisted behavioral sleep intervention (Sleep-Opt-In) was developed and tested to improve sleep duration and regularity in adults with T1D who had either short or irregular sleep patterns. The specific aims of this pilot study were to (1) evaluate the feasibility and acceptability and (2) examine the preliminary outcomes of Sleep-Opt-In on sleep duration and regularity, glucose indices, and patient-reported outcomes (daytime sleepiness, fatigue, diabetes distress, and depressive mood) in adults with T1D and habitual short or irregular sleep.
The Sleep-Opt-In intervention was based on the theoretical framework of supportive accountability.29 Human support is conceptualized as a means to enhance adherence to behavioral technology interventions when the supportive individual is considered legitimate, trustworthy, and benevolent.29 This model suggests that behavioral intervention technology users are more likely to adhere if they are accountable to another person at some time in the future. Accountability is enhanced if goal setting and progress are known to the other person.30 A coaching protocol was designed and tested that was specific to persons with T1D using these principles that had been demonstrated to enhance adherence in a previous sleep intervention.31
A randomized controlled parallel trial design was employed. In this type of design, eligible participants were randomly assigned to either the Sleep-Opt-In intervention or a Healthy Living attention control group. Participants remained in their assigned groups for the remainder of the study. Changes in the outcomes of interest (sleep, glucose, and patient-reported outcomes) were measured at baseline and at the end of the study. Fourteen adults with T1D were recruited and enrolled. The study protocol was approved by the Office for the Protection of Human Subjects, University of Illinois Chicago (Protocol #2018-0762) and was registered with Clinical Trials.gov (NCT# 03617770). Data were collected from January to August 2019.
Interested potential participants were screened for eligibility by telephone by the principal investigator using a screening instrument form. Persons were eligible if they were aged 18 to 65 years, diagnosed with T1D for a minimum of 1 year, and had self-reported sleep duration of less than 6.5 hours per night on workdays or an irregular sleep-wake schedule (defined as greater than 1.5 hours variability in typical bed and/or wake times). Exclusion criteria included conditions that are known to independently affect sleep or glucose, including pregnancy, breastfeeding, severe hypoglycemia in the past 6 months, rotating shift or work hours between 12:00 am and 5:00 am, renal impairment, significant medical comorbidities, and diagnosis or high risk for sleep apnea or insomnia symptoms based on scores on the Berlin Questionnaire (sleep apnea score >2)32 or the Insomnia Sleep Index (insomnia score ≥22).33
This pilot feasibility study was not powered to detect significant effects for improvement in sleep, glycemic control, or diabetes distress. Our sample of 14 participants was determined to be adequate for gaining experience in using this protocol with a T1D population and for calculating mean and variability confidence intervals for sleep and diabetes-related measures.34
Recruitment was conducted through the University of Illinois Chicago Medical Center, diabetes websites, and diabetes organizations using flyers, e-announcements, and recruitment letters. Interested potential participants were contacted by phone by the principal investigator to discuss study procedures and complete assessment for eligibility. If eligible and interested, a study visit was scheduled at the University of Illinois Chicago for baseline data collection and initiation of the study.
Once informed consent was obtained, baseline assessments of sleep, glucose, and patient-reported outcomes (daytime sleepiness, fatigue, diabetes distress, depressive mood) were obtained by trained research assistants over a 1-week run-in phase. The participants were then randomized in the proportion of 3:2 to either the Sleep-Opt-In or Healthy Living attention control group. This randomization scheme permitted more participants to be randomized to the intervention group, to evaluate effect size for a small sample. Randomization was set up by the study statistician using the Research Electronic Data Capture system (REDCap). Study staff who analyzed outcome data were masked to group assignment. The intervention lasted for 8 weeks for both groups. Baseline assessments were repeated at Week 8 (program completion).
Those randomized to Sleep-Opt-In participated in an 8-week remotely delivered program that entailed: (1) weekly emailed didactic sleep content, (2) weekly telephone coaching (5-10 minutes in length), and (3) a wearable sleep tracker and smartphone application with interactive feedback and tools (Fitbit Charge 2®). The didactic sleep content was scheduled and delivered through REDCap. Content was viewable on a smartphone, desktop, or tablet device and took 8 to 10 minutes to read. Once the reading was completed, participants clicked a link to REDCap that provided a timestamp of lesson completion. Information from the lessons was reinforced in the telephone coaching sessions and included topics such as the importance of sleep, sleep procrastination, diabetes, and sleep (Table 1). The wearable sleep tracker allowed participants to track their sleep each day, set goals, and share results with the coach at each weekly coaching phone call. The sleep tracker was used as a simple technology to engage the participants in setting goals and monitoring progress. It was not used to measure sleep outcomes for the study.
All participants were assigned to a trained sleep coach to monitor their progress during the study and provide weekly telephone coaching sessions related to their sleep-related goals. The previously developed coaching protocol was based on the principles of supportive accountability.35 Coaches provided feedback to the participant based on wearable sleeptracker data. At each call, progress was reviewed, problem solving occurred for barriers to progress, and new goals were set for the following week. Between sessions, coaches were available to troubleshoot any problems.
The attention control group was intended to control for the didactic content and coach contact in the intervention group so that we could test the intervention-related components (eg, motivational enhancement, goal setting, feedback, and sleep education). Participants assigned to the control group were provided weekly health education emails requiring 8 to 10 minutes to read (eg, nutrition, stretching exercises). Once the reading was completed, participants clicked a link to REDCap that documented the date and time of lesson completion. They were instructed to maintain their sleep schedule as usual. Participants received weekly brief telephone contact from a different “Healthy Living” coach (ie, 5 min) to answer questions about the lesson content. They did not receive a sleep tracker and did not receive any sleep content.
Feasibility and acceptability. Feasibility was determined through analysis of recruitment, enrollment (number recruited and screened, eligible, and consented), and retention (% session participation, program completion rates). Acceptability of Sleep-Opt-In was determined through written participant evaluation at the end of the intervention. Two components were assessed. (1) Usability of technology was evaluated using the 17-item, 7-point Likert, Modified Usability Scale (MUS) a standard, reliable, and valid measure for this concept.36 (2) Content and structure of the program was evaluated using a 13-item, 7-point Likert scale developed by the investigators. For both scales, higher numbers indicated more positive responses. The Cronbach’s alpha of the MUS scale for this sample was .881.
Objective sleep. Participants wore a sleep actigraph watch (Phillips Actiwatch Spectrum Plus, Respironics, Bend, OR) on their nondominant wrist for 1 week at Weeks 0 and 8 to measure sleep parameters. Actiwatch is a researchgrade, validated, sleep-monitoring device with 96% sensitivity and 86% accuracy compared to the gold-standard polysomnography.37 Participants were instructed to wear the watch for 24 hours per day for 7 days (removing for bathing) and to press an event marker on the watch at bedtime and wake-up time to document sleep time onset and offset. A sleep log was also obtained over the same period in which participants recorded their bedtime and wake-up time as well as any disruptions during the night. Data on the Actiwatch were collected at 30-second epochs with default settings. The data were downloaded at the end of each collection period and reviewed with each participant to clarify any inconsistencies among the watch, event marker, and sleep log. Using the Phillips Actiware software 6.0, rest intervals were manually scored using a standardized scoring algorithm38 and the program-computed sleep duration (total sleep time in minutes), sleep efficiency (% of time in bed spent sleeping), sleep start time, and sleep end time. Midsleep was the time between sleep onset and wake time. The standard deviation of the midsleep time was used to indicate sleep regularity.
Glycemic control was measured using a blood draw for hemoglobin A1C (Quest Diagnostics). This is a standard biomarker for average blood glucose over 3 months.39 Glucose variability was measured using a blinded continuous glucose monitor (CGM; Abbott FreeStyle Libre Professional) that was worn for 1 week. The CGM sensor was placed in the subcutaneous tissue in the upper arm and recorded interstitial glucose levels continuously at 15-minute intervals over the 1-week period. The data were downloaded for analysis. Variables obtained from the CGM included coefficient of variation (CV%) and time in range (the percent time with glucose levels between 70 and 180 mg/dL [3.9-10 mmol/L]).39 Interstitial glucose levels using the Abbott FreeStyle Libre have been found to be accurate and acceptable for insulin dosing decisions (mean absolute difference, 10%).40 Glucose indices were measured at Weeks 0 and 8.
Daytime sleepiness was measured using the Epworth Sleepiness Scale (ESS). The ESS is an 8-item, 4-point Likert scale that evaluates the likeliness of dozing (would never doze to high chance of dozing) during 8 situations. Higher scores indicate greater daytime sleepiness. The scale has been psychometrically validated and differentiates persons with sleep disorders from healthy adults.41 The Cronbach’s alpha for this sample was .627.
Fatigue was measured with the PROMIS Short Form 8a,42 an 8-item, 5-point Likert scale that measures the level of fatigue from tiredness to exhaustion. Response pattern scoring is used, and the results are provided as a t score. The scale has been validated and is indicated for use across all populations.42 The Cronbach’s alpha for this study was .917.
Diabetes distress was measured with the Type 1 Diabetes Distress Scale. This 28-item, 6-point Likert scale measures the emotional and behavioral challenges associated with living with, and managing, T1D.43 Diabetes distress is conceptually distinct from depression.44 The scale is comprised of 7 subscales that represent sources of distress: powerlessness, negative social perceptions, physician distress, friends/family distress, hypoglycemia, management, and eating distress. A total score and subscale scores are obtained. Higher scores indicate greater degrees of diabetes distress. Mean cutoff points have been established: little or no distress (1.0-1.4), low (1.5-1.9), moderate (2.0-2.9), and high (≥3) distress. The scale has been psychometrically validated in persons with T1D.43 The Cronbach’s alpha for the total score for this study was .963.
Depressive mood was measured with the Center for Epidemiologic Studies Depression Scale (CES-D). The CES-D is a 20-item, 4-point Likert-style scale that assesses emotions and behaviors over the past week, providing a single overall depressive symptom score ranging from 0 to 60. A score ≥16 represents the presence of depressive symptoms. The scale has been validated in adult populations.45 The Cronbach’s alpha for this study was .875.
Aim 1. Feasibility and acceptability. Frequencies, percentages, and means (SD) were calculated to demonstrate the feasibility of recruitment, adherence, retention, and acceptability. Qualitative comments were reviewed.
Aim 2. Outcomes of Sleep-Opt-In on sleep duration and regularity, glucose indices, and patient-reported outcomes (daytime sleepiness, fatigue, diabetes distress, depressive mood). Changes in mean pre- and postprogram scores were calculated using descriptive statistics for the key variables in the Sleep-Opt-In and attention control groups. Lastly, for preliminary exploration, a t test for independent samples was used to examine the change in outcomes between groups (SPSS 27).
Forty adults with T1D were screened; 26 were ineligible due to symptoms of sleep apnea, insomnia, or a diagnosis of type 2 diabetes. The remaining 14 were randomized 3:2 to the Sleep-Opt-In or Healthy Living attention control group (ns = 9 and 5, respectively). One participant in the Sleep-Opt-In group withdrew at Week 4 of the intervention; the remaining 13 participants completed the intervention (Figure 1).
The 14 participants were 64% female, 79% White, 21% Latino, with an age range of 19 to 42 years. One half (50%) were married, and 71% had earned a college degree or higher. Mean diabetes duration was 16.5 years. A1C levels ranged from 5.4% to 9.0%; however, 50% met the recommended glycemic target of A1C < 7%. Slightly more than one-third met the time-in-range goal for glucose of 70% or greater39 (Table 2).
Actigraphy-derived sleep measures revealed that there were no significant differences in sleep parameters between the Sleep-Opt-In and control groups at baseline. Sleep duration ranged from 307 to 470 minutes per night, while sleep variability ranged from 21 to 113 minutes over the 7-day period (Table 2). The average time in bed was 491 minutes, while the average sleep duration was 399 minutes. The average sleep efficiency was 81%, and nearly one-third (30%) achieved a sleep efficiency greater than 85% at baseline. The average sleep start time was slightly after midnight (00:13 hours), and average wake time was 07:43 hours. Of note, 6 participants did not demonstrate concordance between self-reported and objectively reported sleep duration and timing. Thus, although 6 participants perceived themselves to have short or irregular sleep, this was not confirmed by actigraphy; sleep duration was within the recommended 7 to 9 hours per night, and sleep variability in bedtime/wake-up time was within 1.5 hours.
Of the 14 eligible participants, all agreed to participate in the study. One participant in the Sleep-Opt-In group withdrew at Week 4 of the intervention (with no reasons provided), representing a 93% completion rate. The remaining participants (n = 13) attended 90% of the weekly lessons and coaching calls and completed the 8-week intervention. All participants in the attention control group completed the program and attended 100% of the sessions. Participants in the Sleep-Opt-In group completed a program evaluation at the end of the intervention. Participants rated the program positively. Usability of the technology (sleep tracker and associated apps) was rated an average of 6.2 on a 7-point scale. Evaluation of the content and structure of the Sleep-Opt-In session was rated an average of 6.1 on a 7-point scale. Comments were also collected, which included, “It made me realize even small improvements in sleep can be beneficial.” “I learned to prioritize my sleep and noticed how it affected my health.” “I liked seeing my sleep stats in real-time on the app.” “Goal-setting with my coach was very beneficial.” “I found that improved sleep habits improved my diabetes control and vice versa.” One suggestion was to “extend the number of weeks of the program,” allowing some additional “gap” weeks between meetings to work on skill development and behavior change.
It was also determined that sleep indices, glucose parameters (A1C, CGM), and patient-reported outcome measures could be obtained successfully with minimal missing actigraphy, CGM, or questionnaire data. One limitation was using self-reported sleep to determine inclusion criteria without actigraphy confirmation. This resulted in 6 participants enrolling in the study who would have been ineligible based on actigraphy recordings.
Change scores from baseline to the end of the 8-week intervention were calculated for the key variables for the Sleep-Opt-In and control groups (Table 3). In the full group of 14 participants, those in the Sleep-Opt-In group demonstrated an improvement in sleep regularity, glycemic variability (CV%), and time in range compared to the attention control group. There was no improvement in sleep duration. Daytime sleepiness, fatigue, diabetes distress, and depressive mood also improved among Sleep-Opt-In participants compared to the attention control group. However, because 6 participants did not have objectively short or irregular sleep, and therefore did not have areas to improve, analysis focused on the subgroup who had documented short or irregular sleep (n = 8).
Objective sleep parameters. Sleep duration improved in both intervention (n = 5) and control groups (n = 3); however, the improvement was greater in the control group (8 minutes vs. 25 minutes, respectively). Sleep regularity improved in the Sleep-Opt-In group (reduced 25 minutes), whereas the control group experienced a negligible change in sleep regularity (2 minutes).
Glucose. A1C was reduced in both Sleep-Opt-In and control groups (0.1% and 0.9%, respectively). Of note, the Sleep-Opt-In group had low A1C levels at baseline compared to the control group (mean 6.6% and 7.16%, respectively), and it would not have been safe to expect lowering of these levels. Glucose variability (CV%) and time in range improved only for Sleep-Opt-In (−3.2% and 6.9%, respectively) and worsened for the control group (0.3% and −5.7%, respectively).
Patient-reported outcomes. Self-reports of daytime sleepiness, fatigue, diabetes distress, and depressive mood improved for Sleep-Opt-In, with worsening outcomes in the control group. The improvements in fatigue and depressive mood were statistically significant (P = .024 and P = .041, respectively; Table 3).
The main objectives of this pilot study were to evaluate the feasibility and acceptability of a sleep intervention (Sleep-Opt-In) targeted for adults with T1D with short or irregular sleep and to examine the effects of Sleep- Opt-In on sleep duration and regularity, glucose indices, and patient-reported outcomes. The findings revealed that Sleep-Opt-In was feasible and acceptable to the target population. Recruitment was successful, and all who qualified consented to enroll. All but 1 participant remained in the study through program completion, reflecting a 93% retention rate. The participants evaluated the Sleep-Opt-In program positively.
Of those with objectively confirmed short or irregular sleep, sleep irregularity improved by 25 minutes on average, whereas sleep duration improved only negligibly (8 minutes). One explanation for this is that most participants at baseline (all except 1) experienced sleep irregularity, not short sleep duration. Sleep coaching was thus focused on setting more consistent bed- and waketimes for these individuals. Improvement in sleep regularity by 25 minutes may be clinically important. In a study by Chontong et al10, a difference in sleep regularity of ≈20 minutes in adults with T1D was associated with ≈0.6% difference in A1C. Zhu et al46 also reported that variability in sleep timing was a significant predictor of A1C in adults with type 2 diabetes.
Of interest, the control group experienced an increase in sleep duration but no change in sleep regularity. The control group was highly reactive, and despite instructions not to change their sleep, they may have extended their sleep duration because they were enrolled in a sleep study. It is interesting to note that the control group might have extended their sleep duration but, without sleep coaching, would not have known to change their sleep timing. This is supported by the data.
The improvement in sleep regularity over the 8-week period may provide an explanation for the improved shortterm glucose indices observed in the Sleep-Opt-In group. Glycemic variability, measured as CV%, decreased from 37.6% to 34.4% (−3.2% change), while time in range increased from 59.4% to 66.3% (+6.9% change). In the control group, these indices remained stable or worsened (respectively). Within-person variability in sleep timing has been linked with circadian misalignment.47 Circadian misalignment has been associated with reduced glucose metabolism in persons with T1D diabetes.10 Much of the literature on sleep health has focused on improving mean sleep duration. Our findings support emerging evidence that shifts in sleep timing are an important pathway to investigate in persons with T1D.
Assessment of glycemic control revealed a reduction of A1C in both groups, with a larger reduction in the control group. A1C is an indicator of the average blood glucose over a 3-month period. In this study, A1C measurement was repeated at 8 weeks and therefore would not fully represent changes due to the intervention. Second, Sleep-Opt-In participants began with an average A1C of 6.6%. This achieves the American Diabetes Association glycemic target of less than 7%; it might be unsafe to reduce A1C below this. Participants in the control group began with an average A1C greater than 7% and could safely reduce their levels.
The improvement in all patient-reported outcomes (daytime sleepiness, fatigue, diabetes distress, and depressive mood) in Sleep-Opt-In was an important finding. It is important to consider that improvement in sleep timing by 25 minutes may have had this effect. The hypothesis was that there would be a reduction in diabetes distress with improved sleep. Sturt et al48 reported that poor sleep was a source of distress among those with both type 1 and type 2 diabetes. In a prior study using focus groups in adults with T1D, sleep time emerged as one of the most challenging times of the day.24 Participants discussed the distress associated with managing nocturnal glucose levels as well as feeling poorly rested after a night of hyperglycemia.24 Fatigue is highly prevalent in persons with T1D (from 26.4% to 40% reported), and sleep disturbances are a modifiable contributor for fatigue.49,50 Reducing fatigue has important clinical, self-management, and quality-of-life implications.51 Mechanistic pathways of sleep and inflammation have been identified in autoimmune disorders and are an area for future research.52
Because this was a pilot study to support future investigation, the sample size was not powered to determine efficacy. In fact, effect sizes are considered imprecise and may not represent an effect obtained with a larger sample.53 Also, as noted, 42% of the sample self-reported short or variable sleep that was not concordant with actigraphy recordings. Among these participants with adequate or regular sleep, there was limited opportunity to improve sleep with our intervention; thus, adding screening with objective sleep measures prior to interventions is recommended in future studies.
Available evidence suggests that sleep impairment reduces glucose control and exacerbates diabetes distress among those with T1D. Sleep impairment can lead to long-term health sequelae, including increased cardiovascular morbidity and mortality. The American Diabetes Association Medical Standards of Care recommend that sleep pattern and duration be assessed in person with diabetes54 but do not provide information on how sleep may be improved. In this pilot study, feasibility of providing a sleep intervention to adults with T1D was evaluated. Sleep- Opt-In was feasible, acceptable, and promising for further evaluation as a means to improve sleep duration or regularity, glucose parameters, and important patient-reported outcomes of diabetes distress, daytime sleepiness, fatigue, and depressive mood in the T1D population.
We wish to acknowledge the editing assistance of Kevin Grandfield, publications manager at UIC Department of Biobehavioral Nursing Science.
This work was supported by the Chicago Center for Diabetes Translation Research (Grant No. NIH/NIDDK P30 DK092949); Dean’s Office of the Biological Sciences, Division of the University of Chicago; and the University of Illinois at Chicago (Grant No. CCTS UL1TR002003), 5R01EY029782.
Pamela Martyn-Nemeth https://orcid.org/0000-0002-0323-3614
From College of Nursing, Department of Biobehavioral Nursing Science, University of Illinois Chicago, Chicago, Illinois (Dr Martyn-Nemeth, Dr Quinn, Ms Chapagai); Department of Psychiatry, College of Medicine, University of Illinois Chicago, Chicago, Illinois (Dr Duffecy); College of Nursing, Department of Population Health Nursing Science, University of Illinois Chicago, Chicago, Illinois (Dr Steffen); Division of Public Health, Department of Family and Preventive Medicine, University of Utah, Salt Lake City, Utah (Dr Baron); Office of Research Facilitation, College of Nursing, University of Illinois Chicago, Chicago, Illinois (Ms Burke); and Division of Endocrinology, Diabetes, and Metabolism, Department of Medicine, University of Illinois Chicago, Chicago, Illinois (Dr Reutrakul).
Corresponding Author:Pamela Martyn-Nemeth, Department of Biobehavioral Nursing Science, College of Nursing, University of Illinois Chicago, 845 S. Damen Ave., #720, M/C 820, Chicago, IL 60612, USA.Email: pmartyn@uic.edu
