Abstract
Study Objectives:
The purpose of this study was to describe objective sleep-wake characteristics and glycemia over 7–14 days in young adults with type 1 diabetes. In addition, person-level associations among objective sleep-wake characteristics (total sleep time, sleep variability, and sleep fragmentation index), daytime sleepiness, and glycemia (glycemic control and glucose variability) were examined.
Methods:
In this cross-sectional study, objective sleep-wake characteristics were measured via actigraphy and glucose variability via continuous glucose monitoring over 6–14 days. At baseline, participants completed the Psychomotor Vigilance Test, the Trail Making Test, and questionnaires on daytime sleepiness, sleep quality, and sleep disturbance including sleep diaries.
Results:
Forty-six participants (mean age, 22.3 ± 3.2 years) wore a wrist actigraph and underwent continuous glucose monitoring concurrently for 6–14 days. Greater sleep variability was directly associated with greater glucose variability (mean of daily differences; r = .33, P = .036). Higher daytime sleepiness was directly associated with greater glucose variability (mean of daily differences; r = .50, P = .001). The association between sleep variability and glucose variability (mean of daily differences) was no longer significant when accounting for daytime sleepiness and controlling for type 1 diabetes duration (P > .05). A higher sleep fragmentation index was associated with greater glucose variability (B = 1.27, P = .010, pr2 = 0.40) after controlling for type 1 diabetes duration and accounting for higher daytime sleepiness.
Conclusions:
Sleep-wake variability, sleep fragmentation, daytime sleepiness, and the associations with glycemia are new dimensions to consider in young adults with type 1 diabetes. Sleep habits in this population may explain higher glucose variability, and optimizing sleep may improve overall diabetes management.
Citation:
Griggs S, Hickman RL Jr, Strohl KP, Redeker NS, Crawford SL, Grey M. Sleep-wake characteristics, daytime sleepiness, and glycemia in young adults with type 1 diabetes. J Clin Sleep Med. 2021;17(9):1865–1874.
Keywords: type 1 diabetes, sleep, psychomotor vigilance, neurocognitive function, self-management, glycemic control, glucose variability
BRIEF SUMMARY
Current Knowledge/Study Rationale: Young adults have the poorest glycemic control of any age group with type 1 diabetes and are at high risk of developing complications as they navigate multiple transitions while assuming responsibility for their diabetes self-management. However, research to date has been based on self-reported sleep measures, short measurement time frames, and intermittent glucose monitoring to characterize sleep and glycemia across a broad age range of adults with type 1 diabetes.
Study Impact: Shorter sleep leads to poorer glycemia and diabetes self-management, but the results of this study emphasize the connection between daytime sleepiness, sleep fragmentation, sleep variability, and greater glucose variability in young adults with type 1 diabetes. There is a need to address sleep-wake behaviors to foster better diabetes self-management in this high-risk population.
INTRODUCTION
Type 1 diabetes (T1D), a catabolic autoimmune disorder arising from the destruction of insulin-producing beta cells, is one of the most common chronic conditions in young adulthood, affecting 1.6 million Americans.1 In the United States, only 14%–30% of young adults ages 18–30 years with T1D achieve targets for glycemic control (glycosylated hemoglobin [A1C] < 7.0%).2 Poorer glycemic control and greater glycemic variability (fluctuations outside of range) are associated with an increased risk of both micro- and macrovascular complications, and up to 50% of young adults develop diabetes-related complications in their 20s.2 Compared to age- and sex-matched peers, young adults with T1D perform poorer on neuropsychological tests of psychomotor speed, sustained attention, and speed of information processing.3,4 Such findings are typically interpreted as a disease-related outcome of T1D (eg, chronic hyperglycemia, microangiopathy, age of onset, or disease duration)3,4; however, there may be modifiable behaviors such as sleep-wake times that may explain this difference.
Short sleep duration (< 6.5 hours per night) is associated with hypoglycemia,5,6 hyperglycemia,7,8 and poorer glycemic control in young adults with T1D.7–9 Young adults with T1D are at a higher risk for short sleep duration because of interactions among diabetes self-management practices, biology (eg, later chronotype, delayed melatonin secretion), and environment (eg, early school/work wake-up times, light exposure, social interactions, transition from pediatric to adult diabetes care).4,5 Better glycemic control is associated with better overall sleep quality7 and a higher percentage of time spent in deep sleep measured by polysomnography in adults with T1D.8 Sleep variability (day-to-day variability in total sleep time) is associated with poorer glycemic control in adolescents10 and middle-aged adults (40–65 years) with T1D11,12 who are younger and older than the current sample (18–30 years). Young adults with T1D have a more pronounced sleep extension on weekends compared to those without chronic conditions7; however, whether there are relationships among sleep variability and glycemia or sleep-wake behaviors and neurocognitive function in young adults with T1D has not been explored. Thus, it is essential to identify sleep-wake behaviors as possible modifiable barriers or facilitators to achieving glycemic control in this population.
Studies to date have used cross-sectional descriptive designs (n < 30) across a broad range of ages with short measurement time frames (eg, over 3 days), self-reported sleep measures, and/or intermittent blood glucose monitoring rather than using current monitoring technologies (eg, continuous glucose monitors [CGM]).7,13,14 The gold standard for determining glycemic control in research and practice is measured via A1C; however, A1C does not account for glycemic variability or hypoglycemia in those with T1D. These glycemic fluctuations contribute to endothelial damage and the onset or progression of premature complications.15 We addressed this by monitoring both A1C and CGM in the current study. In addition, weaknesses in previous study designs are addressed in the current study by objectively monitoring sleep-wake and glucose concurrently over 7–14 days to reflect habitual sleep-wake and glucose patterns with the additional ability to assess weekend and weekday differences.
Thus, the primary purpose of this cross-sectional study was to characterize sleep and glycemia over 7–14 days among young adults ages 18–30 years with T1D. In addition, person-level (between-persons) associations among objective sleep-wake characteristics (total sleep time, variability in total sleep time [sleep variability], sleep fragmentation index), daytime sleepiness, and glycemia (glycemic control and glucose variability) were explored. The hypothesis was that poorer objective sleep-wake characteristics (shorter total sleep time, more sleep variability, and higher sleep fragmentation index), and higher daytime sleepiness were associated with poorer glycemia (poorer glycemic control and more glucose variability).
METHODS
Methods have been reported in a previously published protocol paper,16 and details pertinent to this study are described in the sections below.
Participants
Young adults with T1D were recruited from the Yale New Haven Health System (New Haven, CT) who were (1) between ages 18 and 30 years, (2) diagnosed with T1D for at least 6 months, (3) diagnosed with no other major health problems (eg, chronic medical conditions or major psychiatric illness), (4) not participating in any intervention studies, and (5) readers/speakers of English. Those with a previous obstructive sleep apnea diagnosis, those who were night shift workers, and those currently pregnant were excluded. This age range was chosen because individuals at this key developmental stage must navigate multiple transitions (eg, education, occupation, living situation, moving from pediatric to adult diabetes care) while also assuming responsibility for their diabetes self-management. The Berlin Questionnaire was used to screen participants for inclusion in the study.17 Participants considered to be at high risk for sleep apnea were referred for treatment and not included in the study.
A variety of recruitment methods were used. Participants were recruited from MyChart (84.8%), face to face in the clinic (8.7%), or through a flyer/verbal referral from clinic providers (6.5%). Out of 450 potential participants identified in the system as meeting the criteria through MyChart, 18 declined, 65 read the message with no response, 91 expressed interest, 57 were determined eligible after chart or phone screening, and 46 consented and completed the study (Figure 1). Those who declined to participate did so because of the time commitment of the study required by the protocol, a lack of interest, or the refusal/inability to wear a CGM. Participants were recruited over a period of 14 months from December 2018 to February 2020. All study participants (n = 46, 100%) had actigraphy data ranging from 6–14 days/nights (mean, 8.7 ± 2.6 days/nights) and a majority (n = 43, 93.5%) had CGM data ranging from 6–14 days/nights (mean, 9.2 ± 3.9 days/nights).
Figure 1. Study screening.
Procedure
Institutional review board approval was obtained from both Yale University (New Haven, CT; 2000023502) and Case Western Reserve University (Cleveland, OH; 20200650). After written informed consent was obtained, young adults completed questionnaires, a 10-minute psychomotor vigilance test, and a 2-minute test of executive function (Trail Making Tests [TMT] A and B).18 The participants were given the Phillips Respironics Spectrum Plus (Koninklijke Philips, Nevada), a wrist-worn device, to wear continuously on their nondominant wrist for 14 days. To demarcate time in bed, participants were instructed to depress the event marker at “lights out” and “lights on.”19 We scored actigraph rest intervals for bedtime and wake time from a combination of the event marker, diary, and light level. Priority was given for the event marker and light level (agreement for 2 out of 3) because these events are in real time. The young adults who did not already have a CGM were instructed to insert the Dexcom G4 sensor (Dexcom Inc, San Diego, California) during the initial study visit and to wear it continuously for 2 weeks with the Spectrum Plus. There was a 2-hour calibration period for the provided CGM, and participants were instructed to calibrate the receiver by entering 2 separate fingerstick glucose values from their personal glucometer into the receiver when prompted after the study visit. The principal investigator (PI) was on call for any questions that arose after the study visit, and participants were provided with 24/7 Dexcom support services. Participants also needed to calibrate the CGM twice daily during the study period with their personal glucometer.
REDCap is a secure web-based software that was used to administer the questionnaires at baseline with twice-daily diaries for the monitoring period (7–14 days). Each diary entry was time-stamped, so it was evident when participants completed each survey to reduce the potential for recall bias. Electronic medical record data were entered directly into REDCap. Young adults completed sleep diaries daily in the mornings and evenings to track daytime sleep-related behaviors (eg, caffeine use, exercise) and nocturnal sleep characteristics (eg, bedtime, awakenings). The PI communicated with participants via call, text, or email based on their preference the day after enrollment to address any problems and again at the end of the week to remind them to complete the surveys and return the watch and diary in a prepaid mailer. The PI tracked the completion of the twice-daily diaries and the upload of the CGM report at the end of the week on the REDCap dashboard and sent reminders via email when missing data were noted. Young adults received a small incentive ($25) for their time to complete questionnaires and received additional incentives for returning the watch ($35) and CGM (if applicable) and for participating in an exit interview ($40).
Variables and measures
Demographic and clinical characteristics
A combination of interviews and the electronic medical record was used to collect clinical and demographic data, including age, body mass index (kg/m2), duration of diabetes, most recent A1C, and medical history (diabetes, sleep, last menstrual period, and other). The following data were collected via a self-report survey: ethnicity, education, primary caregiver, employment status, full-time student status, work hours, marital status, residence, household count, income, cigarette smoking, alcohol or other substance use, insulin therapy regimen (eg, insulin injections or insulin pump), CGM pump brand (if applicable), and last menstrual period. Self-report data were also validated through the electronic medical record.
Self-reported sleep characteristics
At baseline, participants self-reported sleep characteristics, sleep disturbance, sleepiness, and total sleep time. Participants recorded wake time, bedtime, sleep, and symptoms twice daily for 14 days while wearing the wrist actigraph.20 Participants logged the timing and frequency of awakenings and bed/rise times in the Pittsburgh Sleep Diary and related etiology on a visual analog scale (fatigue, nocturia, pain, hypo/hyperglycemia). Global sleep quality was reported using the 18-item Pittsburgh Sleep Quality Index (Cronbach’s α = 0.87).21 Pittsburgh Sleep Quality Index component scores are summed and range from 0–21, with higher scores indicating poorer sleep quality.22 The Cronbach’s α for the Pittsburgh Sleep Quality Index in the current study was 0.75. Participants rated their sleep disturbance on the 8-item PROMIS Sleep Disturbance scale (Cronbach’s α = 0.90 or above), with higher scores indicating greater sleep disturbance.23 The Cronbach’s α for the PROMIS Sleep Disturbance scale in the current study was 0.90.
Daytime sleepiness
Participants’ general level of daytime sleepiness was measured with the 8-item Epworth Sleepiness Scale (Cronbach’s α = 0.88).24 Scores range from 0–24, with higher scores being indicative of more sleepiness.24 The Cronbach’s α for the Epworth Sleepiness Scale in the current study was 0.64.
Objective sleep-wake characteristics
Participants were instructed to wear the Spectrum Plus for 14 days. The Spectrum Plus collects activity data with a standard spectrum of light and off-wrist detection. Five days or longer of monitoring reduces inherent measurement errors and increases reliability.25 Actigraphy provides a measure of general sleep-wake characteristics including total sleep time, sleep efficiency (%), wake after sleep onset, sleep onset latency, and sleep fragmentation index (movement index + fragmentation index). Actigraphy provides the greatest agreement and the least bias in comparison with polysomnography in young adults with T1D for all sleep-wake parameters except for sleep onset latency (P = .04).26
Neurocognitive function
The Psychomotor Vigilance Test (PVT), a 10-minute test administered by an investigator on a PVT-192 device (CWE, Inc., Ardmore, PA), measures simple reaction time to stimuli occurring at random intervals and also measures vigilant attention. Sleep loss, extended wakefulness, circadian misalignment, and time on task is assessed using the PVT.27 The following measures were derived from the PVT: (1) mean response time (RT), (2) median RT, (3) the 10% slowest RT, (4) the 10% fastest RT, and (5) the number of lapses (defined as RT > 500 milliseconds—ie, the inability to respond in a timely fashion when a stimulus is present). PVT performance does not improve with repeated administration.27 The PVT has high reliability, with intraclass correlations measuring test-retest reliability above 0.80 in young adults.27
The PI (SG) administered the TMT, a 3- to 5-minute test on paper consisting of parts A and B.28 The TMT provides information on visual search, scanning, speed of processing, mental flexibility, and executive functions. For the TMT-A, an individual must connect lines sequentially for 25 encircled numbers, and for the TMT-B, an individual must alternate between numbers and letters (eg, 1, A, 2, B, 3, C). Seventy-five seconds is the adult average for the TMT-B, with deficiencies noted at > 273 seconds.28 The TMT has high reliability and has been validated with young adults with a test-retest reliability of TMT-A of 0.76 to 0.89 and TMT-B of 0.86 to 0.94, respectively.29 Both parts were used in this study. This measure was scored by subtracting the score of the TMT-A from that of the TMT-B. Higher scores reflect poorer executive function, and the score represents the amount of time required to complete the task. The Cronbach’s α for the TMT in the current study was 0.71.
Diabetes self-management
The Self-Care Inventory-Revised is a 15-item questionnaire used to assess a person’s ability to perform diabetes self-management behaviors (diet, exercise, blood glucose monitoring, insulin administration, and attending medical appointments).30 The Self-Care Inventory-Revised was validated with U.S. and Dutch patients with T1D and has very good internal consistency (Cronbach’s α = 0.87). Scores range from 15–75, with higher scores reflecting better diabetes self-management.30 The Cronbach’s α in the current study was 0.57.
Glycemia
Glycemic control was measured by the most recent A1C level that was routinely measured at quarterly clinic visits using the Siemens Vantage Glucose Analyzer (range, 2.5%–14%; Arsie et al).31 Glucose variability was determined from the CGM data that were downloaded directly from each participant’s existing CGM or the provided blinded Dexcom G4 CGM to capture glucose patterns. CGM systems provide real-time, dynamic glucose information every 5 minutes—up to 288 readings in a 24-hour period.32 Participants used an automatic inserter to insert a small sensor wire just under their skin.32 CGMs are accurate across a wide range of levels, with test-retest reliability ranging from 0.77–0.95,33 diagnostic sensitivity 78%, and specificity 96%.34 Glucose variability indices were calculated from the CGM as mean ± standard deviation, time in range (calculated as the percentage in the target range, 70–180 mg/dL; J index, the overall quality of glucose variability [calculated as 0.001 × (mean + SD)2]), low and high blood glucose indices, mean amplitude of glucose excursion (average amplitude of upstrokes or downstrokes with magnitude > 1 standard deviation), and mean of daily differences.35
Statistical analyses
The PI managed data on the REDCap site and exported all data into SPSS version 26 (SPSS for Mac, IBM Corp., Armonk, NY) and SAS 9.4 (SAS Institute, Inc., Cary, NC) for analysis. Actigraphy data were scored with Actiware version 6.0.9 software. CGM data were calculated with Glyculator version 2.0 software (Koninklijke Philips, Nevada).36 PVT data were scored with REACT software for PVT data analysis.
A quantitative descriptive approach was used to characterize sleep-wake, neurocognitive function, diabetes self-management, glycemic control, and glucose variability among the 46 young adults with T1D over the 7–14 days to capture weekend and weekday differences. For these analyses, objective sleep-wake characteristics were averaged across the days of monitoring. Glucose variability indices were calculated based on data across the days. At least 6 days of actigraphy data were required to characterize sleep reliably.25 Descriptive statistics were used to summarize each of the self-report and objective variables. Objective sleep-wake characteristics were described including total sleep time, variability in total sleep time, sleep onset latency, sleep efficiency, wake after sleep onset, and sleep fragmentation index over the 2-week interval.
Bivariate correlations were used to explore the relationships among total sleep time, sleep variability, sleep fragmentation, daytime sleepiness, glycemic control, and glucose variability. Descriptive statistics were used to summarize each of the variables, including the scores for multi-item scales. Using actigraphy data, sleep variability was calculated using the mean square of successive differences of total sleep time in addition to the participant-specific standard deviation of total sleep time across the 14 nights, representing the within-persons variation in sleep night to night. These approaches for analyzing sleep variability have been documented in prior research.10,37 A1C was used for glycemic control, and CGM data were used to calculate glucose variability indices.35
To evaluate explanatory contributions of sleep-wake characteristics to glucose variability (mean of daily differences), we performed a series of linear regression models for each of the sleep-wake variables that were statistically significant in the unadjusted associations. In the first model of each analysis, we included the significant actigraphy-derived sleep characteristic with the covariate T1D duration. In the second model, we included daytime sleepiness to evaluate its independent effect.
RESULTS
Sample characteristics
Demographic and clinical characteristics are presented in Table 1. Participants were mostly non-Hispanic White, (84.8%) and female (67.4%) and had T1D for 10.3 years (± 6.0 years). The majority used an insulin pump for treatment (80.4%) and CGM for monitoring (87%). The mean A1C was 7.2% (± 1.1%) with 56.5% (n = 26) of participants achieving glycemic targets (A1C < 7%), suggesting that glycemic control was, on average, acceptable. Mean glucose measured via CGM was 163 mg/dL (± 30.2 mg/dL). Actigraphy and glycemic control data were available for all participants (n = 46), and CGM data were available for a majority of participants (n = 43, 93.5%). For CGM systems, participants used the Dexcom G6 (54.8%, n = 23), the Dexcom G5 (19.0%, n = 8), the Medtronic (Medtronic MimniMed, CA) MiniMed 670G (11.9%, n = 5), and the Freestyle Libre (Abbott Diabetes Care Inc. Alameda, CA) (2.4%, n = 1). Participants without a CGM were provided with a blinded Dexcom G4 (11.9%, n = 5).
Table 1.
Sample characteristics (n = 46).
| Characteristics | |
|---|---|
| Age (y) | 22.3 ± 3.2 |
| Sex (% male) | 15 (32.6) |
| BMI | 27.0 ± 4.4 |
| Race | |
| White | 39 (84.8) |
| Asian | 3 (6.5) |
| Black or African American | 1 (2.2) |
| Native Hawaiian or other Pacific Islander | 1 (2.2) |
| Unknown | 1 (2.2) |
| Ethnicity (% Hispanic or Latino) | 7 (15.2) |
| Annual income (USD) | |
| < 9,999 | 3 (6.5) |
| 10,000–19,999 | 1 (2.2) |
| 20,000–49,999 | 2 (4.3) |
| 50,000–99,999 | 9 (19.6) |
| 100,000–149,999 | 6 (13.0) |
| > 150,000 | 10 (21.7) |
| Unsure | 13 (28.3) |
| Choose not to answer | 2 (4.3) |
| Monthly expenses (% able to cover) | 43 (93.5) |
| Education | |
| Senior in high school | 1 (2.2) |
| High school graduate | 4 (8.7) |
| Some college, no degree | 18 (39.1) |
| Associate degree | 1 (2.2) |
| Bachelors degree | 19 (41.3) |
| Masters degree | 3 (6.5) |
| Marital status | |
| Never married | 40 (87.0) |
| Married | 6 (13.0) |
| Full-time college student (yes) | 25 (54.3) |
| Work hours | |
| Not working | 11 (23.9) |
| 2-14 h/wk | 7 (15.0) |
| Part-time (15–35 h/wk) | 11 (23.9) |
| Full-time (≥35 h/wk) | 17 (37.0) |
| Residence | |
| Owned | 9 (19.6) |
| Rented | 13 (28.3) |
| Live with friends | 1 (2.2) |
| Live with family | 22 (47.8) |
| On campus | 1 (2.2) |
| Household count | 3.8 ± 1.3 |
| Smoking (yes) | 1 (2.2) |
| Alcohol use (AUDIT-C) | 2.9 ± 1.96 |
| Type 1 diabetes profile | |
| T1D duration (y) | 10.3 ± 6.0 |
| A1C (%) | 7.2 ± 1.1 |
| Glucose mean (CGM) | 163.0 ± 30.2 |
| Insulin pump (% yes) | 37 (80.4) |
| CGM pump brand | |
| Dexcom G6 | 24 (52.2) |
| Dexcom G5 | 8 (17.4) |
| Medtronic Guardian | 7 (15.2) |
| Freestyle Libre | 1 (2.2) |
| N/A, does not use CGM | 6 (13.0) |
For continuous variables, normally distributed data are presented as mean ± SD. Data for categorical variables are presented as n (%). A1C = glycosylated hemoglobin, AUDIT C = alcohol use disorders identification Test, BMI = body mass index, CGM = continuous glucose monitor, SD = standard deviation, T1D = type 1 diabetes, USD = US dollars.
Self-reported sleep characteristics
Descriptive statistics for sleep-wake characteristics, neurocognitive function, self-management, and glucose variability indices are presented in Table 2. Participants reported a mean Pittsburgh Sleep Quality Index score of 5.91 (± 3.5). The mean sleep disturbance was reported as 45.25 (± 8.36), which was slightly better than that of peers without diabetes.38
Table 2.
Descriptive statistics for sleep-wake characteristics, neurocognitive function, self-management, and glucose variability indices.
| Measure | n | Mean or (%) | ± SD |
|---|---|---|---|
| TST (min) | 46 | 421.53 | 64.37 |
| Sleep onset latency (min) | 46 | 19.23 | 13.14 |
| Sleep efficiency (%) | 46 | 85.02 | 4.80 |
| Wake after sleep onset (min) | 46 | 36.66 | 17.61 |
| Sleep fragmentation index (% + %) | 46 | 17.82 | 5.67 |
| Sleep variability (min), SD of TST | 46 | 70.32 | 30.10 |
| Sleep variability, MSSD of TST | 46 | 10,886.0 | 11,519.8 |
| Sleep-wake questionnaires | |||
| Overall sleep quality (PSQI) | 46 | 5.91 | 3.5 |
| Sleep disturbance (PROMIS) | 46 | 45.29 | 8.36 |
| Sleepiness (ESS) | 46 | 7.52 | 3.29 |
| Neurocognitive function | |||
| Cognitive function (PROMIS) | 46 | 51.35 | 8.08 |
| TMT score (B-A) | 45 | 18.68 | 8.91 |
| Psychomotor vigilance (PVT) | |||
| Mean response time (RT) | 46 | 300.34 | 110.14 |
| Median RT | 46 | 276.95 | 96.73 |
| Mean RT, slowest 10% | 46 | 2.43 | 0.65 |
| Mean RT, fastest 10% | 46 | 210.98 | 46.73 |
| Number of RT > 500 ms lapses | 46 | 5.10 | 15.0 |
| Diabetes self-management | |||
| Self-Care Inventory | 46 | 39.80 | 6.20 |
| Glucose variability | |||
| % time in range 70–180 mg/dL | 43 | 61.70 | 15.66 |
| % time in range 70–140 mg/dL | 43 | 39.43 | 14.70 |
| J index | 43 | 51.79 | 19.80 |
| High blood glucose index | 43 | 8.18 | 4.66 |
| Low blood glucose index | 43 | 1.06 | 1.36 |
| MAGE | 42 | 151.87 | 39.74 |
| MODD | 42 | 64.74 | 21.19 |
The sleep fragmentation index is the movement index plus the fragmentation index. It includes both restlessness and fragmentation of the sleep period. ESS = Epworth Sleepiness Scale, MAGE = mean amplitude of glucose excursion, MODD = mean of daily differences, MSSD = mean square of successive differences, PSQI = Pittsburgh Sleep Quality Index, PVT = Psychomotor Vigilance Test, RT = response time, SD = standard deviation, TMT = Trail Making Test, TST = total sleep time.
Objective sleep-wake characteristics
Across all days/nights of actigraphy data, the mean total sleep time ranged from 5 hours 24 minutes–9 hours 25 minutes, mean sleep onset latency ranged from 2.86 minutes–59.07 minutes, mean sleep efficiency ranged from 72.97%–92.75%, mean wake after sleep onset ranged from 0 minutes–72.07 minutes, and mean sleep fragmentation indices ranged from 7.7–40.13. Only 45.7% (n = 21) of participants habitually achieved > 7 hours of actigraphy-measured sleep on average.39,40 There were no significant differences in objective sleep-wake characteristics or glycemia by age, sex, body mass index, or T1D duration. Participants slept more on the weekends (mean difference, 15.3 minutes); however, the difference between weekdays and weekends was not statistically significant (P = .19). Full details of the actigraphy sleep scoring were previously reported.16
Neurocognitive function and diabetes self-management
As measured by the PVT, the mean RT ranged from 219.2–895.9 milliseconds, the median RT ranged from 213.0–786.0 milliseconds, the 10% slowest RT ranged from 0.6–3.9, the 10% fastest RT ranged from 148.3–433.0 milliseconds, and the number of lapses ranged from 0–83. Participants in the present study had a shorter mean and median RT compared to those reported in the literature for peers without diabetes (300.3 vs 325.0 milliseconds and 276.9 vs 278.0 milliseconds).41 However, participants in the present study had more lapses on average compared to peers without diabetes (5.1 vs 0.39) or those with insufficient sleep syndrome (5.1 vs 4.3).42
TMT-A scores ranged from 11.6–36.6 (mean, 18.1 ± 5.9), TMT-B scores ranged from 18.1–82.1 (mean, 36.7 ± 12.0), and TMT total scores ranged from 5.1–47.2. Participants in this study had better executive function than same-age peers without diabetes (TMT, 18.1 ± 5.9 vs TMT-A, 22.9 ± 6.9 and TMT-B, 36.7 ± 12.0 vs 49.0 ± 12.7, respectively).18
Self-reported sleepiness ranged from 2–16, and self-reported cognitive function ranged from 35.6–47.2. Mean sleepiness and self-reported cognitive function scores were consistent with those of same-age peers (7.6 ± 3.9 vs 7.6 ± 3.324 and 51.3 vs 50.0).38 Self-reported diabetes self-management scores ranged from 23–52. Participants in the present study had poorer mean diabetes self-management scores than same-age peers with T1D (39.8 ± 6.2 vs 57.9 ± 10.6).30
Objective sleep-wake characteristics, daytime sleepiness, and glycemia
Bivariate analyses were conducted to determine the associations among objective sleep characteristics, daytime sleepiness, and glycemia (Table 3). Higher sleep variability was directly associated with greater glucose variability (mean of daily differences; r = .33, P = .036). Higher daytime sleepiness was directly associated with greater glucose variability (mean of daily differences; r = .50, P = .001). The contributions of daytime sleepiness with sleep variability and the sleep fragmentation index in separate models to glucose variability after controlling for T1D duration are presented in Table 4 and Table 5.
Table 3.
Bivariate correlations between total sleep time, sleep variability, daytime sleepiness, and glycemia (glycemic control and glucose variability indices).
| Total Sleep Time r (P) | Sleep Variability r (P) | Sleep Fragmentation Index r (P) | Daytime Sleepiness r (P) | |
|---|---|---|---|---|
| Glycemic control (A1C) | 0.08 (.621) | 0.17 (.170) | 0.17 (.258) | 0.16 (.275) |
| Time in range 70–180 mg/dLa | –0.20 (.209) | –0.07 (.641) | –0.19 (.223) | –0.19 (.218) |
| J indexa | 0.09 (.585) | 0.16 (.322) | 0.24 (.127) | 0.22 (.166) |
| Mean of daily differencesa | 0.01 (.935) | 0.33 (.036)* | 0.27 (.076) | 0.50 (.001)* |
aCGM variables. *P < .05. A1C = glycosylated hemoglobin, CGM = continuous glucose monitor.
Table 4.
Contributions of sleep variability and daytime sleepiness to MODD (regression model).
| Model 1 | Model 2 | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| B | SE | β | P | Pr2 | B | SE | β | P | Pr2 | |
| Predictors | ||||||||||
| Sleep variability | 0.00 | 0.00 | 0.32 | .057 | 0.30 | 0.00 | 0.00 | 0.10 | .089 | 0.10 |
| Daytime sleepiness | 3.0 | 1.0 | 0.45 | .007 | 0.42 | |||||
| Covariate | ||||||||||
| T1D duration | –0.11 | 0.56 | –0.03 | .849 | –0.03 | –0.39 | 0.63 | –0.07 | .664 | –0.07 |
| R2 | .106 | .265 | ||||||||
| F2 | .199 | .667 | ||||||||
B = unstandardized coefficient regression coefficient, β = standardized regression coefficient, F2 = Cohen’s F effect size, MODD = mean of daily differences, Pr2 = partial correlation coefficient shown for each variable in each model, R2 = coefficient of determination shown for each model, SE = standard error, T1D = type 1 diabetes.
Table 5.
Contributions of sleep variability and daytime sleepiness to MODD (regression model).
| Model 1 | Model 2 | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| B | SE | β | P | Pr2 | B | SE | β | P | Pr2 | |
| Predictors | ||||||||||
| Sleep fragmentation index | 0.96 | 0.56 | 0.27 | .093 | 0.27 | 1.3 | 0.47 | 0.35 | .010 | 0.40 |
| Daytime sleepiness | 3.7 | 0.87 | 0.55 | < .001 | 0.56 | |||||
| Covariate | ||||||||||
| T1D duration | –0.37 | 0.54 | –0.11 | .500 | –0.11 | –0.18 | 0.45 | –0.05 | .697 | –0.06 |
| R2 | .087 | .377 | ||||||||
| F2 | .245 | .488 | ||||||||
B = unstandardized coefficient regression coefficient, β = standardized regression coefficient, F2 = Cohen’s F effect size, MODD = mean of daily differences, Pr2 = partial correlation coefficient shown for each variable in each model, R2 = coefficient of determination shown for each model, SE = standard error, T1D = type 1 diabetes.
In the first model, we examined sleep variability (Table 4). The association between sleep variability and the mean of daily differences was no longer significant when T1D duration was added to the model (Table 4). In the second model, we examined the unique contribution of daytime sleepiness and higher daytime sleepiness was associated with a greater mean of daily differences (P = .005, pr2 = .42), accounting for 26.5% of the variance (Cohen’s F = .667). The association between daytime sleepiness and sleep variability was no longer significant when daytime sleepiness was added to the model (P = .089; Table 4).
In the next set of models, we examined the sleep fragmentation index (Table 5). In the first model, the association between the sleep fragmentation index and the mean of daily differences was no longer significant when T1D was added to the model. In the second model, both higher daytime sleepiness (P < .001, pr2 = .56) and a higher sleep fragmentation index (P = .010, sr2 = .40) were associated with greater glucose variability (mean of daily differences), accounting for 37.7% of the variance (Cohen’s F = .488; Table 6). In separate models with body mass index and sex, there was not a significant change in R2 (data not shown) for either sleep variability or the sleep fragmentation index, so T1D duration was determined to be the covariate with the best model fit given the variability in T1D in the sample (10.3 ± 6.0 years).
Table 6.
Contributions of sleep fragmentation index and daytime sleepiness to glycemic control and glucose variability indices (regression model).
| Outcome Variables | ||||
|---|---|---|---|---|
| Glycemic Control (A1C), R2 = .065 | J Index, R2 = .120 | MODD, R2 = .377 | Time in Range 70–180 mg/dL, R2 = .084 | |
| Predictors | ||||
| Sleep fragmentation index | 0.04 ± 0.03 (.198) | 0.89 ± 0.52 (.095) | 1.27 ± 0.47 (.010)* | –0.59 ± 0.42 (.168) |
| Daytime sleepiness | 0.06 ± 0.05 (.210) | 1.49 ± 0.93 (.117) | 3.66 ± 0.87 (.000)* | –1.07 ± 0.75 (.163) |
| Covariate | ||||
| T1D duration | 0.00 ± 0.03 (.940) | –0.17 ± 0.50 (.730) | –0.18 ± 0.45 (.697) | –0.03 ± 0.40 (.949) |
Values are presented as unstandardized coefficients ± standard error (P value). *Values are significant at P < .05. A1C = glycosylated hemoglobin, MODD = mean of daily differences, T1D = type 1 diabetes.
DISCUSSION
The hypothesis about shorter total sleep time and glycemic control/glucose variability was not supported in the current study; however, sleep variability, daytime sleepiness, and sleep fragmentation were confirmed as important factors in the associations with greater glucose variability. Sleep variability, not shorter total sleep time, was directly associated with greater glucose variability in this population. The majority of the participants in the sample exhibited high variability in total sleep time and poor self-reported sleep quality.
Shorter total sleep time was not directly associated with glucose variability, glycemic control, or the other variables of interest. The association between shorter total sleep time and poorer glycemic control has varied in other studies. Findings in the present study were consistent with other studies of adults and adolescents with T1D.7,43,44 However, shorter total sleep time was associated with poorer glycemic control in 2 previous studies of adults with T1D.13,14 The methods in those studies were different, with researchers in 1 study relying on intermittent glucometer testing to determine glucose variability,7 and researchers in other studies focusing on self-reported sleep duration.43,44 Glycemic control measured via A1C is retrospective and may not reflect current trends in glucose data.45
On the other hand, higher sleep variability was directly associated with daytime sleepiness and greater glucose variability (mean of daily differences). Higher sleep variability likely reflects sleep deprivation alternating with sleep compensation along with shifts in circadian timing. The young adults in the present study slept slightly but not significantly longer on the weekend compared to weekdays. This finding was consistent with previous studies of adolescents10,46 and adults11 with T1D. Sleep variability was no longer significant, and it was daytime sleepiness that accounted for the variance in greater glucose variability (mean of daily differences) in the model. However, both sleep fragmentation and daytime sleepiness accounted for variance in glucose variability in the young adults in the current study.
The associations were significant only between sleep fragmentation, sleep variability, and daytime sleepiness and 1 glucose variability index (mean of daily differences), and not the other indices (time in range or J index). There are several methods to quantify glucose variability and no universally accepted gold standard; however, most often, interday variation (mean of daily differences) best reflects the swings of blood glucose in a person with diabetes as a consequence of an absent autoregulation and the shortcomings of insulin therapy (ie, T1D and the required intensive insulin management).47 Thus, it is important to examine different aspects of glucose variability, such as the interday variation, that reflect the mean absolute value of the differences between glucose values on 2 consecutive days at the same time (ie, the mean of daily differences).48 Measuring glucose variability with mean of daily differences is distinct from other indices such as the J index, which distinguishes the overall quality of variability, excluding severe and persistent hypoglycemia; the time in range (percentage of time in the target range 70-180 mg/dL); and the time spent in a condition of hypoglycemia (< 54 mg/dL).48
The findings emphasize a need for the concurrent objective characterization of sleep and glycemia over a longer period of time in a narrower age range of young adults with T1D with impact on neurocognitive function. In previous studies, a broader range of ages has been included, and this range may lead to an underestimation of the associations between total sleep time and glycemic control or glucose variability. The range of days that participants wore the actigraph was 6–14 days, with a mean of 8.7 ± 2.6 days/nights, and the continuous glucose monitor period was 6–14 days/nights, with a mean of 9.2 ± 3.9 days/nights. Although this duration may not fully reflect habitual sleep and glucose patterns, current recommendations for actigraphy suggest that 72 hours is adequate for monitoring sleep.25 The rigorous objective neurocognitive measures used in this study allowed for the capture of aspects of objectively measured fatigue that are often not picked up with self-reported measures.
These results should be considered in the context of the study’s limitations. The present study sample was mostly White (84.8%); therefore, racial/ethnic differences in key variables of interest could not be determined. Participants in the present sample were mostly socioeconomically advantaged (93.5%) and female (67.4%), with slightly better glycemic control (43.5% compared to 30% nationally) and higher CGM use (87% compared to 30% nationally), and from 1 geographic location, limiting generalizability. In addition, the design was cross-sectional; therefore, the direction of the associations cannot be inferred. The lack of significance in the associations among total sleep time, sleep variability, and glycemic control may result from the relatively small number of participants.
Data on physical activity and lifestyle including diet, insulin, or other hormones (eg, growth hormone, cortisol) were not collected. Increased glucose levels are seen with increased levels of growth hormone and cortisol.49 Growth hormone, secreted in deep stage N3 sleep, reduces insulin sensitivity.50 A strength of this study is that those with a previous obstructive sleep apnea diagnosis and those with a high risk for sleep apnea were screened for exclusion, reducing the risk for an independent impact of obstructive sleep apnea on glycemic control. However, laboratory polysomnography was not used, so there may still have been participants with sleep apnea in this sample. Other sleep disorders were not evident but were not systemically excluded. Despite not controlling for these other factors in the present study, the findings are novel, and associations over a longer period of time than in previous studies were found.7,13,14 Future studies where hormonal, dietary, and insulin treatment effects are controlled and monitored over a longer period of time capturing more than 1 or 2 weekends may provide further insight into the findings presented here.
Clinicians should routinely assess sleep habits and aim to optimize sleep in this high-risk population. Researchers should continue to investigate the mechanistic pathways between sleep and glucose variability at the day level (within-person measurements) to better understand temporal ordering. This approach would help provide insight into whether poor sleep precedes greater glucose variability or vice versa, or whether it acts bidirectionally. A focus on sleep is appropriate because it is a modifiable target that may improve neurocognitive function, glycemic control, and glucose variability. How to implement sleep extension and sleep time stability may be a challenge in this population because of the intensive self-management regimen associated with T1D. Future research should address these gaps.
DISCLOSURE STATEMENT
All authors have seen and approved the final version of this manuscript. S.G. is funded by the National Institute for Nursing Research (K99NR018886) and the American Academy of Sleep Medicine (220-BS-19). S.G. was previously funded by the National Institute for Nursing Research (T32 NR0008346) and Sigma Theta Tau International provided research support for the study. The contents of this article are solely the responsibility of the authors and do not necessarily represent the official view of the National Institutes of Health or the American Academy of Sleep Medicine. Dexcom provided continuous glucose monitors (G4) free of charge to be used in the study for participants who did not have their own device. The remaining authors report no conflicts of interest.
ACKNOWLEDGMENTS
Author contributions: S.G. designed the study; collected, analyzed, and interpreted the data; and drafted the manuscript. S.C. contributed to the study design, analyzed and interpreted the data, and drafted the manuscript. R.L.H. contributed to the study design, analyzed and interpreted the findings, and drafted the manuscript. M.G. and K.P.S. contributed to the study design, reviewed the content, and revised the manuscript. N.S.R. contributed to the study design and reviewed the content.
ABBREVIATIONS
- A1C
glycosylated hemoglobin
- CGM
continuous glucose monitor
- PI
principal investigator
- PVT
Psychomotor Vigilance Test
- RT
response time
- TMT
Trail Making Test
- T1D
type 1 diabetes
REFERENCES
- 1. Introduction: standards of medical care in diabetes—2020 . Diabetes Care. 2020. ; 43 ( Suppl 1 ): S1 – S2 . [DOI] [PubMed] [Google Scholar]
- 2. Bryden KS , Dunger DB , Mayou RA , Peveler RC , Neil HA . Poor prognosis of young adults with type 1 diabetes: a longitudinal study . Diabetes Care. 2003. ; 26 ( 4 ): 1052 – 1057 . [DOI] [PubMed] [Google Scholar]
- 3. Inkster BE , Zammitt NN , Ritchie SJ , Deary IJ , Morrison I , Frier BM . Effects of sleep deprivation on hypoglycemia-induced cognitive impairment and recovery in adults with type 1 diabetes . Diabetes Care. 2016. ; 39 ( 5 ): 750 – 756 . [DOI] [PubMed] [Google Scholar]
- 4. Jauch-Chara K , Hallschmid M , Schmid SM , et al . Plasma glucagon decreases during night-time sleep in type 1 diabetic patients and healthy control subjects . Diabet Med. 2007. ; 24 ( 6 ): 684 – 687 . [DOI] [PubMed] [Google Scholar]
- 5. Banarer S , Cryer PE . Sleep-related hypoglycemia-associated autonomic failure in type 1 diabetes: reduced awakening from sleep during hypoglycemia . Diabetes. 2003. ; 52 ( 5 ): 1195 – 1203 . [DOI] [PubMed] [Google Scholar]
- 6. Borel AL , Benhamou PY , Baguet JP , et al . Short sleep duration is associated with a blood pressure nondipping pattern in type 1 diabetes: the DIAPASOM study . Diabetes Care. 2009. ; 32 ( 9 ): 1713 – 1715 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Barone MTU , Wey D , Schorr F , et al . Sleep and glycemic control in type 1 diabetes . Arch Endocrinol Metab. 2015. ; 59 ( 1 ): 71 – 78 . [DOI] [PubMed] [Google Scholar]
- 8. Feupe SF , Frias PF , Mednick SC , McDevitt EA , Heintzman ND . Nocturnal continuous glucose and sleep stage data in adults with type 1 diabetes in real-world conditions . J Diabetes Sci Technol. 2013. ; 7 ( 5 ): 1337 – 1345 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Denic-Roberts H , Costacou T , Orchard TJ . Subjective sleep disturbances and glycemic control in adults with long-standing type 1 diabetes: the Pittsburgh’s Epidemiology of Diabetes Complications study . Diabetes Res Clin Pract. 2016. ; 119 : 1 – 12 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Patel NJ , Savin KL , Kahanda SN , et al . Sleep habits in adolescents with type 1 diabetes: Variability in sleep duration linked with glycemic control . Pediatr Diabetes. 2018. ; 19 ( 6 ): 1100 – 1106 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Chontong S , Saetung S , Reutrakul S . Higher sleep variability is associated with poorer glycaemic control in patients with type 1 diabetes . J Sleep Res. 2016. ; 25 ( 4 ): 438 – 444 . [DOI] [PubMed] [Google Scholar]
- 12. Larcher S , Gauchez AS , Lablanche S , Pépin JL , Benhamou PY , Borel AL . Impact of sleep behavior on glycemic control in type 1 diabetes: the role of social jetlag . Eur J Endocrinol. 2016. ; 175 ( 5 ): 411 – 419 . [DOI] [PubMed] [Google Scholar]
- 13. Borel AL , Pépin JL , Nasse L , Baguet JP , Netter S , Benhamou PY . Short sleep duration measured by wrist actimetry is associated with deteriorated glycemic control in type 1 diabetes . Diabetes Care. 2013. ; 36 ( 10 ): 2902 – 2908 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Matejko B , Skupien J , Mrozińska S , et al . Factors associated with glycemic control in adult type 1 diabetes patients treated with insulin pump therapy . Endocrine. 2015. ; 48 ( 1 ): 164 – 169 . [DOI] [PubMed] [Google Scholar]
- 15. Torimoto K , Okada Y , Mori H , Tanaka Y . Relationship between fluctuations in glucose levels measured by continuous glucose monitoring and vascular endothelial dysfunction in type 2 diabetes mellitus . Cardiovasc Diabetol. 2013. ; 12 ( 1 ): 1 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Griggs S , Redeker NS , Crawford SL , Grey M . Sleep, self-management, neurocognitive function, and glycemia in emerging adults with type 1 diabetes mellitus: a research protocol . Res Nurs Health. 2020. ; 43 ( 4 ): 317 – 328 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Netzer NC , Stoohs RA , Netzer CM , Clark K , Strohl KP . Using the Berlin Questionnaire to identify patients at risk for the sleep apnea syndrome . Ann Intern Med. 1999. ; 131 ( 7 ): 485 – 491 . [DOI] [PubMed] [Google Scholar]
- 18. Tombaugh TN . Trail Making Test A and B: normative data stratified by age and education . Arch Clin Neuropsychol. 2004. ; 19 ( 2 ): 203 – 214 . [DOI] [PubMed] [Google Scholar]
- 19. Ancoli-Israel S , Martin JL , Blackwell T , et al . The SBSM guide to actigraphy monitoring: clinical and research applications . Behav Sleep Med. 2015. ; 13 ( Suppl 1) : S4 – S38 . [DOI] [PubMed] [Google Scholar]
- 20. Monk TH , Reynolds CF , Kupfer DJ , et al . The Pittsburgh sleep diary . J Sleep Res. 1994. ; 3 ( 2 ): 111 – 120 . [PubMed] [Google Scholar]
- 21. Backhaus J , Junghanns K , Broocks A , Riemann D , Hohagen F . Test-retest reliability and validity of the Pittsburgh Sleep Quality Index in primary insomnia . J Psychosom Res. 2002. ; 53 ( 3 ): 737 – 740 . [DOI] [PubMed] [Google Scholar]
- 22. Buysse DJ , Reynolds CF III , Monk TH , Berman SR , Kupfer DJ . The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research . Psychiatry Res. 1989. ; 28 ( 2 ): 193 – 213 . [DOI] [PubMed] [Google Scholar]
- 23. Yu L , Buysse DJ , Germain A , et al . Development of short forms from the PROMIS™ Sleep Disturbance and Sleep-Related Impairment item banks . Behav Sleep Med. 2011. ; 10 ( 1 ): 6 – 24 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Johns MW . Reliability and factor analysis of the Epworth Sleepiness Scale . Sleep. 1992. ; 15 ( 4 ): 376 – 381 . [DOI] [PubMed] [Google Scholar]
- 25. Sadeh A . The role and validity of actigraphy in sleep medicine: an update . Sleep Med Rev. 2011. ; 15 ( 4 ): 259 – 267 . [DOI] [PubMed] [Google Scholar]
- 26. Farabi SS , Quinn L , Carley DW . Validity of actigraphy in measurement of sleep in young adults with type 1 diabetes . J Clin Sleep Med. 2017. ; 13 ( 5 ): 669 – 674 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Dorrian J , Rogers NL , Dinges DF . Psychomotor vigilance performance: neurocognitive assay sensitive to sleep loss . In: Kushida CA , ed. Sleep Deprivation: Clinical Issues, Pharmacology, and Sleep Loss Effects. New York, NY: : Marcel Dekker; ; 2004. ; 39 – 70 . [Google Scholar]
- 28. Gaudino EA , Geisler MW , Squires NK . Construct validity in the Trail Making Test: what makes Part B harder? J Clin Exp Neuropsychol. 1995. ; 17 ( 4 ): 529 – 535 . [DOI] [PubMed] [Google Scholar]
- 29. Wagner S , Helmreich I , Dahmen N , Lieb K , Tadić A . Reliability of three alternate forms of the Trail Making Tests A and B . Arch Clin Neuropsychol. 2011. ; 26 ( 4 ): 314 – 321 . [DOI] [PubMed] [Google Scholar]
- 30. Weinger K , Butler HA , Welch GW , La Greca AM . Measuring diabetes self-care: a psychometric analysis of the Self-Care Inventory-Revised with adults . Diabetes Care. 2005. ; 28 ( 6 ): 1346 – 1352 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Arsie MP , Marchioro L , Lapolla A , et al . Evaluation of diagnostic reliability of DCA 2000 for rapid and simple monitoring of HbA1c . Acta Diabetol. 2000. ; 37 ( 1 ): 1 – 7 . [DOI] [PubMed] [Google Scholar]
- 32. Wagner J , Tennen H , Wolpert H . Continuous glucose monitoring: a review for behavioral researchers . Psychosom Med. 2012. ; 74 ( 4 ): 356 – 365 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Terada T , Loehr S , Guigard E , et al . Test-retest reliability of a continuous glucose monitoring system in individuals with type 2 diabetes . Diabetes Technol Ther. 2014. ; 16 ( 8 ): 491 – 498 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Jensen MH , Christensen TF , Tarnow L , Mahmoudi Z , Johansen MD , Hejlesen OK . Professional continuous glucose monitoring in subjects with type 1 diabetes: retrospective hypoglycemia detection . J Diabetes Sci Technol. 2013. ; 7 ( 1 ): 135 – 143 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Suh S , Kim JH . Glycemic variability: how do we measure it and why is it important? Diabetes Metab J. 2015. ; 39 ( 4 ): 273 – 282 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Czerwoniuk D , Fendler W , Walenciak L , Mlynarski W . GlyCulator: a glycemic variability calculation tool for continuous glucose monitoring data . J Diabetes Sci Technol. 2011. ; 5 ( 2 ): 447 – 451 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Meltzer LJ , Sanchez-Ortuno MJ , Edinger JD , Avis KT . Sleep patterns, sleep instability, and health related quality of life in parents of ventilator-assisted children . J Clin Sleep Med. 2015. ; 11 ( 3 ): 251 – 258 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Cella D , Riley W , Stone A , et al . PROMIS Cooperative Group . The Patient-Reported Outcomes Measurement Information System (PROMIS) developed and tested its first wave of adult self-reported health outcome item banks: 2005-2008 . J Clin Epidemiol. 2010. ; 63 ( 11 ): 1179 – 1194 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Hirshkowitz M , Whiton K , Albert SM , et al . National Sleep Foundation’s sleep time duration recommendations: methodology and results summary . Sleep Health. 2015. ; 1 ( 1 ): 40 – 43 . [DOI] [PubMed] [Google Scholar]
- 40. Watson NF , Badr MS , Belenky G , et al . Consensus Conference Panel . Joint Consensus Statement of the American Academy of Sleep Medicine and Sleep Research Society on the recommended amount of sleep for a healthy adult: methodology and discussion . J Clin Sleep Med. 2015. ; 11 ( 8 ): 931 – 952 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Zhang C , Varvarigou V , Parks PD , et al . Psychomotor vigilance testing of professional drivers in the occupational health clinic: a potential objective screen for daytime sleepiness . J Occup Environ Med. 2012. ; 54 ( 3 ): 296 – 302 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Thomann J , Baumann CR , Landolt HP , Werth E . Psychomotor Vigilance Task demonstrates impaired vigilance in disorders with excessive daytime sleepiness . J Clin Sleep Med. 2014. ; 10 ( 9 ): 1019 – 1024 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. van Dijk M , Donga E , van Dijk JG , et al . Disturbed subjective sleep characteristics in adult patients with long-standing type 1 diabetes mellitus . Diabetologia. 2011. ; 54 ( 8 ): 1967 – 1976 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Yeshayahu Y , Mahmud FH . Altered sleep patterns in adolescents with type 1 diabetes: implications for insulin regimen . Diabetes Care. 2010. ; 33 ( 11 ): e142 . [DOI] [PubMed] [Google Scholar]
- 45. Pickup JC , Freeman SC , Sutton AJ . Glycaemic control in type 1 diabetes during real time continuous glucose monitoring compared with self monitoring of blood glucose: meta-analysis of randomised controlled trials using individual patient data . BMJ. 2011. ; 343 : d3805 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Rechenberg K , Griggs S , Jeon S , Redeker N , Yaggi HK , Grey M . Sleep and glycemia in youth with type 1 diabetes . J Pediatr Health Care. 2020. ; 34 ( 4 ): 315 – 324 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Siegelaar SE , Holleman F , Hoekstra JB , DeVries JH . Glucose variability: does it matter? Endocr Rev. 2010. ; 31 ( 2 ): 171 – 182 . [DOI] [PubMed] [Google Scholar]
- 48. Wilmot EG , Choudhary P , Leelarathna L , Baxter M . Glycaemic variability: the under-recognized therapeutic target in type 1 diabetes care . Diabetes Obes Metab. 2019. ; 21 ( 12 ): 2599 – 2608 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49. Leproult R , Van Cauter E . Role of sleep and sleep loss in hormonal release and metabolism . Endocr Dev. 2010. ; 17 : 11 – 21 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50. Davidson JR , Moldofsky H , Lue FA . Growth hormone and cortisol secretion in relation to sleep and wakefulness . J Psychiatry Neurosci. 1991. ; 16 ( 2 ): 96 – 102 . [PMC free article] [PubMed] [Google Scholar]