A Comparison of the Rates of Clock-Based Nocturnal Hypoglycemia and Hypoglycemia While Asleep Among People Living with Diabetes: Findings from the Hypo-METRICS Study

Abstract

Introduction:

Nocturnal hypoglycemia is generally calculated between 00:00 and 06:00. However, those hours may not accurately reflect sleeping patterns and it is unknown whether this leads to bias. We therefore compared hypoglycemia rates while asleep with those of clock-based nocturnal hypoglycemia in adults with type 1 diabetes (T1D) or insulin-treated type 2 diabetes (T2D).

Methods:

Participants from the Hypo-METRICS study wore a blinded continuous glucose monitor and a Fitbit Charge 4 activity monitor for 10 weeks. They recorded details of episodes of hypoglycemia using a smartphone app. Sensor-detected hypoglycemia (SDH) and person-reported hypoglycemia (PRH) were categorized as nocturnal (00:00–06:00 h) versus diurnal and while asleep versus awake defined by Fitbit sleeping intervals. Paired-sample Wilcoxon tests were used to examine the differences in hypoglycemia rates.

Results:

A total of 574 participants [47% T1D, 45% women, 89% white, median (interquartile range) age 56 (45–66) years, and hemoglobin A1c 7.3% (6.8–8.0)] were included. Median sleep duration was 6.1 h (5.2–6.8), bedtime and waking time ∼23:30 and 07:30, respectively. There were higher median weekly rates of SDH and PRH while asleep than clock-based nocturnal SDH and PRH among people with T1D, especially for SDH <70 mg/dL (1.7 vs. 1.4, P < 0.001). Higher weekly rates of SDH while asleep than nocturnal SDH were found among people with T2D, especially for SDH <70 mg/dL (0.8 vs. 0.7, P < 0.001).

Conclusion:

Using 00:00 to 06:00 as a proxy for sleeping hours may underestimate hypoglycemia while asleep. Future hypoglycemia research should consider the use of sleep trackers to record sleep and reflect hypoglycemia while asleep more accurately.

The trial registration number is NCT04304963.

Introduction

Hypoglycemia is a common adverse effect of insulin treatment in people with diabetes.¹ During sleep, physiological responses to hypoglycemia and mental arousal are reduced, leading to a lower likelihood that individuals will wake up and take corrective action.^2
–4 As a result, hypoglycemic episodes occurring at night tend to last longer and be more severe than those occurring during waking hours.⁴ In addition, nocturnal hypoglycemia has been associated with lower quality of life, higher fatigue levels, and reduced productivity.⁵ Qualitative evidence has also shown that people with diabetes report fear and anxiety about going to sleep at night, often linked with greater worries about hypoglycemia and fear of dying as a result of hypoglycemia.^6,7

Key research studies to date that have reported on nocturnal hypoglycemia rates^8
–10 used a clock-based method to distinguish between daytime and nighttime, based on the assumption that sleep occurs between predefined hours (e.g., 00:00–06:00 or 23:00–07:00). This approach is pragmatic and designed to provide consistency between studies, but it may have contributed to biased results. This bias is due to the fact that there is intra- and interindividual variability in sleeping patterns in the general population and in people with diabetes, often as a result of biological (e.g., individual's chronotype) and environmental (e.g., employment status or social obligations) factors.^11,12 In addition, given the reduced physiological and behavioral responses to hypoglycemia both relate to sleep status rather than time of day, the ideal would be to measure sleep adequately when estimating the rates of nocturnal hypoglycemia.

Smartphone technologies and wearable devices that capture and monitor daily-life sleep and wake cycles in real time are increasingly available, accessible, and user-friendly.¹³ New-generation devices generally estimate sleep start and end times and sleep stages using sensors, which detect body movement, heart rate, and heart rate variability.¹⁴ Recent validation studies of wearable devices such as the Fitbit (Fitbit, Inc., San Francisco, CA, USA) have shown promising results when compared with the gold standard polysomnography method in accurately measuring sleep in clinical^15,16 and nonclinical populations.¹⁷ While some hypoglycemia studies have used activity monitors to record sleep periods previously,^18,19 no studies to date have assessed whether the rates of conventionally calculated clock-based nocturnal hypoglycemia accurately reflect those of hypoglycemia while asleep. This study therefore aimed to address this gap by comparing the rates of clock-based nocturnal hypoglycemia to those of hypoglycemia while asleep, as measured using a real-time sleep tracker device.

Methods

Study population

Hypoglycaemia—MEasurement, ThResholds and ImpaCtS (Hypo-METRICS) is a 10-week observational study conducted across nine United Kingdom (UK) and European Union (EU) sites between October 2020 and August 2022, as part of the Hypo-RESOLVE project.²⁰ Full details of the study methods have been reported previously.²¹ Participants recruited to the study were 602 adults aged between 18 and 85 years with type 1 diabetes (T1D) or insulin-treated type 2 diabetes (T2D) who had at least one episode of hypoglycemia in the 3 months before the study. People who worked regular night shifts were excluded from the study.

Participants wore a blinded modified continuous glucose monitor (CGM) device (Abbott Freestyle Libre 2™) and a sleep and activity tracking device (Fitbit Charge 4™) continuously for 10 weeks. Participants were asked to record all episodes of hypoglycemia using daily questionnaires and a real-time symptom tracker in the novel Hypo-METRICS smartphone application. Full details of the design of the Hypo-METRICS app have been published previously.²²

Ethical approval for the study was granted by the Oxford B Research Ethics Committee (United Kingdom), CMO Region Arnhem-Nijmegen (Netherlands), Ethikkommission der Medizinischen Universität Graz (Austria), Videnskabsetisk Komite for Region Hovedstaden (Denmark), and Comite De Protection Des Personnes SUD Mediterrannee IV (France). All participants provided informed written consent for the study.

Study sample

Hypo-METRICS participants were included in the present study if Fitbit sleep, CGM and Hypo-METRICS app data were all adequately collected during the study. Participants with no Fitbit and/or CGM data (due to technical issues) were excluded. Participants for whom all hypoglycemic events recorded had missing sleep data at the time of the event were also excluded.

Hypoglycemia classification

Two types of hypoglycemic events were considered: sensor-detected hypoglycemia (SDH) recorded with the CGM, and person-reported hypoglycemia (PRH) recorded with the Hypo-METRICS app. SDH was defined as glucose readings below a threshold (70 and 54 mg/dL) for a minimum of 15 min, in accordance with consensus guidelines.²³ PRH was defined as a symptomatic event that resolved on carbohydrate ingestion or a self-measured glucose below 4 mmol/L, reported by participants using the Hypo-METRICS app.

Night and sleep classification

Nighttime was defined using the hours between 00:00 and 06:00. Sleep time was determined using start time and end time of sleep as generated by the Fitbit Charge 4™, which uses movement patterns and heart rate variability to determine whether a person is asleep.¹⁷

Weekly rates of clock-based nocturnal hypoglycemia and hypoglycemia while asleep

Each hypoglycemic event was categorized based on their clock-based nocturnal status and Fitbit-detected sleep status. An SDH was classified as “nocturnal” if it occurred between 00:00 and 06:00 and “diurnal” if it occurred between 06:01 to 23:59. An SDH was classified as “while asleep” if it occurred between start time and end time of sleep (including daytime sleep) as identified by the Fitbit and “while awake” if it occurred outside sleep intervals. If an SDH interval overlapped between nocturnal/diurnal and/or asleep/awake periods, it was attributed the nocturnal/sleep status for which the area under the curve was the greatest. Each PRH was categorized using the same method as for SDH.

For each participant and hypoglycemic event type (SDH <70 mg/dL, SDH <54 mg/dL and PRH), we calculated 2 weekly rates: the rates of clock-based nocturnal hypoglycemia and the rates of hypoglycemia while asleep, as the number of events divided by the number of study weeks.

Statistical analysis

A descriptive analysis was conducted to describe participants' baseline characteristics. Sleep characteristics as derived from the Fitbit were also described. Categorical variables were presented as frequencies and percentages. Continuous variables were checked for normality using the Shapiro–Wilk test and skewed data were presented as median and interquartile range (IQR). Sleeping patterns were visually assessed first using individual data from two participants as illustrative examples, and second, histograms of the distributions of bedtimes (centered around 00:00) and wake-up times (centered around 06:00) in the overall sample.

Paired-sample Wilcoxon tests were used to estimate the differences in the weekly rates between hypoglycemia while asleep and clock-based nocturnal hypoglycemia. This was conducted for SDH <70 mg/dL, SDH <54 mg/dL, and PRH separately for participants with T1D and T2D.

All analyses were conducted in R (R V.4.2.2 and R Studio V. 2023.06.2 for Windows).

Results

A total of 602 participants were recruited for the Hypo-METRICS study; 589 had sleep data, CGM values, and Hypo-METRICS app data, and of those, 575 were eligible for inclusion in the present study (Fig. 1).

FIG. 1.

Flowchart of study participation. CGM, continuous glucose monitor; PRH, person-reported hypoglycemia; SDH, sensor-detected hypoglycemia.

Participants' characteristics

The study sample included participants who were predominantly white with an almost equal representation of both women and men. Diabetes duration was moderate and glucose control relatively good. One-fifth of the sample had impaired awareness of hypoglycemia. Participants with T1D predominantly used CGM as their usual method of glucose monitoring, while participants with T2D mostly self-monitored their blood glucose (Table 1). Median percentage of time the blinded CGM was active was 95.3% (89.2–98.2) and Hypo-METRICS app questionnaire completion rates 91.1% (84.4–95.7). Almost all participants experienced at least one episode of SDH <70 mg/dL (96.3%), SDH <54 mg/dL (83.9%), or PRH (97.2%) during the study.

Table 1.

Participants' Baseline Characteristics

Characteristics	Overall (N = 575)	Type 1 Diabetes (N = 271)	Type 2 Diabetes (N = 304)
Age (years), median (IQR)	56 (45–66)	47 (30–56.5)	63 (55–69)
HbA_1c (%), median (IQR)	7.3 (6.8–8)	7.3 (6.7–7.8)	7.4 (6.8–8.3)
Diabetes duration (years), median (IQR)	19 (11–27)	19 (9–33)	19 (13–25)
Impaired awareness of hypoglycemia, n (%)	138 (24)	58 (21.4)	80 (26.3)
Sex, n (%)
Women	260 (45.2)	147 (54.2)	113 (37.2)
Men	313 (54.4)	122 (45)	191 (62.8)
Other	2 (0.3)	2 (0.7)	0
Ethnicity, n (%)
Asian	16 (2.8)	3 (1.1)	13 (4.3)
Black	14 (2.4)	4 (1.5)	10 (3.3)
White	512 (89)	238 (87.8)	274 (90.1)
Other	33 (5.7)	26 (9.6)	7 (2.3)
Method of glucose monitoring, n (%)
Continuous glucose monitoring	331 (57.6)	206 (76)	125 (41.1)
Capillary blood glucose monitoring	244 (42.4)	65 (24)	179 (59.1)
Employment status, n (%)
Full-time education	18 (3.1)	15 (5.5)	3 (1)
Full-time employment	216 (37.6)	134 (49.4)	82 (27)
Part-time employment	82 (14.3)	53 (19.6)	29 (9.5)
Retired	205 (35.7)	47 (17.3)	158 (52)
Unemployed	54 (9.4)	22 (8.1)	32 (10.5)

HbA_1c, hemoglobin A1c; IQR, interquartile range.

Sleep characteristics and patterns

Median (IQR) number of nights with sleep data available per participant was 69 (65–70). Participants with T1D had a median time asleep of 6.4 h (5.8–7.0) and participants with T2D had a median time asleep of 5.6 h (4.6–6.5).

In the overall sample, 61.7% of the bedtimes recorded by the Fitbit were between 12:00 and 23:59, and 78.6% of waking-up times were between 06:01 and 18:00. Only about 5% of the sleeping periods recorded by the Fitbit during the study fell within the 00:00–06:00 time window. Figure 2 highlights the intra- and interindividual variability in sleeping patterns in two participants over the course of the study. Figure 3 shows that in participants with T1D, the median bedtime (IQR) [23:28 (22:34; 00:33)] was before 00:00 and wake-up time [07:31 (06:31; 08:35)] after 06:00. Similarly, in participants with T2D, median bedtime was 23:27 (22:08; 01:05) and wake-up time was 07:24 (05:58; 08:49).

FIG. 2.

Sleeping patterns during the study in (A) a participant with type 1 diabetes and (B) a participant with type 2 diabetes. Blue circles indicate bedtimes, green circles indicate getting up times and horizontal lines join up circles corresponding to a unique sleep cycle. For illustration purposes, the time frame has been limited to the period from 20:00 and 13:00 and to sleep cycles lasting over 5 hours.

FIG. 3.

Histograms of the distribution of bedtimes and waking-up times in (A) type 1 diabetes and type 2 diabetes. Thick black lines indicate the median time when going to bed or when getting up. Thick red lines indicate time references used to calculate clock-based nocturnal hypoglycemia rates: 00:00 hours for time when going to bed and 06:00 hours for time when getting up.

Clock-based nocturnal hypoglycemia and hypoglycemia while asleep

Overall, 25,813 SDH <70 mg/dL (30.1% nocturnal and 34.7% while asleep), 6022 SDH <54 mg/dL (39.8% nocturnal and 46.1% while asleep), and 16,254 PRH (20.7% nocturnal and 23.1% while asleep) were recorded. The analysis of the overlap between hypoglycemia while asleep and clock-based nocturnal hypoglycemia events showed that for all hypoglycemia types (SDH <70 mg/dL, SDH <54 mg/dL, and PRH), ∼25%–30% of hypoglycemia events while asleep were not identified as clock-based nocturnal hypoglycemia.

Among participants with T1D, weekly hypoglycemia rates while asleep were higher than those of clock-based nocturnal hypoglycemia for SDH <70 mg/dL, SDH <54 mg/dL, and PRH (all with P < 0.001; Fig. 4A). Among participants with T2D, weekly hypoglycemia rates while asleep were higher than those of clock-based nocturnal hypoglycemia for SDH <70 mg/dL (P < 0.001) and SDH <54 mg/dL (P < 0.001), but not for PRH (Fig. 4B).

FIG. 4.

Weekly rates of clock-based nocturnal hypoglycemia and hypoglycemia while asleep in (A) type 1 diabetes and (B) type 2 diabetes. Thick lines indicate the median, box edges represent the 0.25 and 0.75 quartile range and the black vertical lines the overall range. ***: p < 0.001 PRH, person-reported hypoglycemia; SDH, sensor-detected hypoglycemia.

Discussion

This study is the first to date to compare the rates of hypoglycemia while asleep using a sleep-tracking wearable device with those of clock-based nocturnal hypoglycemia. First, we have shown in both T1D and T2D that median sleep duration was close to 6 h during the study as recorded by the Fitbit. Second, we found that the period from 00:00 to 06:00, typically used to define the night in counting hypoglycemic episodes, did not accurately reflect participants' sleeping patterns. Lastly, the rates of hypoglycemia while asleep were higher than those of clock-based nocturnal hypoglycemia, for SDH <70 mg/dL, SDH <54 mg/dL, and PRH in T1D, and for SDH <70 and <54 mg/dL in T2D.

It was shown that participants in the present study slept for ∼6 h, which is consistent with previous studies reporting sleep durations of 5.97 h,²⁴ 6.2 h,²⁵ 6.3 h,²⁶ and 6.7 h²⁷ measured using accelerometers or wireless sleep monitors in people with diabetes. Nevertheless, the hours between 00:00 and 06:00 did not align with their chronotypes as median bedtime was close to 23:30 and waking-up time close to 7:30 in both T1D and T2D. Comparable sleeping times have been described before in T1D (bedtime: 23:44 and wake-up time: 07:28)²⁸ and T2D (bedtime: 23:30),²⁷ among people categorized as having an “intermediate” chronotype, that is, a person with neither a preference for morning or evening in the sleep and awake cycles.²⁹

Beyond participants' chronotypes, it is also possible that professional, personal, and social obligations, which are part of daily life,^11,12 contributed to the fact that 00:00 to 06:00 did not align with daily-life sleeping patterns in our study. In addition, recruitment for the Hypo-METRICS study was partly conducted during the COVID-19 pandemic, and while a recent study showed that COVID restrictions were not associated with differences in sleep duration in people with T2D in the United Kingdom,³⁰ it is likely that different restrictions over the five countries involved in the present study may have impacted daily lives, including sleeping patterns and timings.

This study has also demonstrated that using a clock-based method to examine nocturnal hypoglycemia underestimates the rates of hypoglycemia while asleep and this may partly be because 00:00 to 06:00 did not align with participants' sleeping patterns, as previously discussed. This partly corroborates data from the Diabetes Control and Complications Trial, which showed that 55% of severe hypoglycemic events started during sleep (episodes of hypoglycemia and their characteristics such as sleep status were prospectively recorded by participants) but only 43% occurred between 00:00 and 08:00 h.³¹ The rates of nocturnal hypoglycemia occurring between 00:00 and 06:00 observed in the present study were higher than previously reported^8,9,32 but still up to 30% lower than those of hypoglycemia while asleep, whether sensor-detected or person-reported.

This is essential as underestimating the rates of hypoglycemia while asleep may mean that we are currently underestimating its impact on factors such as work productivity and general well-being and quality of life for people with diabetes. It would thus be important for future research to establish whether the impact of clock-based nocturnal hypoglycemia differs from that of hypoglycemia while asleep.

Taken together, our findings stress the importance of using real-time wearable sleep trackers when the incidence of hypoglycemia, and particularly hypoglycemia during sleep, is an important research outcome. It is acknowledged that this may not always be feasible, but other methods such as smartphone-based sleep trackers and self-reported sleeping times in diaries could be used, although they may be less accurate than wearable trackers.^33
–35 Nevertheless, these methods have the advantage of recording actual sleeping patterns and timings rather than arbitrarily choosing sleeping times as per the clock-based method, which, as shown in the present study, makes a significant difference to the estimated hypoglycemia rates while asleep.

The main strength of this study is the use of a validated device to objectively record sleep to examine sleeping patterns and estimate the rates of hypoglycemia while asleep, which has never been conducted before in a substantial sample of people with diabetes (n = 575) and for an extensive study duration (n = 10 weeks). In addition, our results are based on robust data as participants' data completion rates were relatively high for the CGM, Fitbit, and Hypo-METRICS app. Our study was limited by the Fitbit Charge device tendency to overestimate time asleep as suggested by previous validation studies.^16,17 Beyond differences in the way sleep/nighttime was defined, it is also possible that factors such as chronotype, short periods of time awake during the night, and percentage of time in each sleep stage, which have not been accounted for in this study, could have contributed to our results. Future research could therefore investigate to which extent those factors contribute to differences between hypoglycemia while asleep and clock-based nocturnal hypoglycemia. We cannot rule out a potential selection bias as recruited participants reported at least one episode of hypoglycemia in the months before the study, which may have inflated hypoglycemia rates in comparison with previous large population-based studies. Nevertheless, the demography of our recruited participants was not otherwise dissimilar from other populations, in terms of age, diabetes duration, or prevalence of impaired hypoglycemia awareness. Lastly, the predominant ethnic group was white in the present study, which may limit the generalizability of the findings.

To conclude, the conventional clock-based method to examine nocturnal hypoglycemia underestimates the rates of hypoglycemia while asleep by up to 30%. Our findings will need to be reproduced but nevertheless suggest that using real-time sleep trackers to record sleep when examining hypoglycemia while asleep may allow researchers and clinicians to (1) better reflect its rates and (2) better account for underpinning biological and environmental factors that can affect sleep duration and timings. Thus, with the explosion of sleep-tracking technologies available, future hypoglycemia research should consider using validated real-time sleep trackers as a standard to report on the rates of hypoglycemia while asleep. When the use of real-time sleep trackers is not feasible, our novel data could allow researchers to estimate the possible error in nocturnal hypoglycemia rates.

Footnotes

Acknowledgments

The authors would like to thank the people with diabetes who participated in the Hypo-METRICS study and the site personnel involved in participant recruitment at each of the clinical centers. The authors also thank Abbott Diabetes Care for providing the continuous glucose monitors used in the study and uMOTIF Limited for providing the platform for the Hypo-METRICS app. Lastly, the authors would like to thank the Hypo-RESOLVE Patient Advisory Committee for their support in the development of the Hypo-METRICS study.

Authors' Contributions

Conceptualization: G.M.-E., P.D., N.Z., S.A., and P.C. Data curation: G.M.-E. and M.G. Formal analysis: G.M.-E. and P.D. Funding acquisition: B.d.G., U.P.-B., R.J.M., E.R, S.H., M.E., J.K.M., S.A.A., J.S., F.P., and P.C. Investigation: G.M.-E. and P.D. Methodology: G.M.-E., P.D., N.Z., S.A., and P.C. Project administration: G.M.-E., P.D., N.Z., U.S., M.B., B.d.G., U.P.-B., R.J.M., E.R., S.H., M.E., J.K.M., S.A.A., J.S., F.P., and P.C. Resources: B.d.G., U.P.-B., R.J.M., E.R., S.H., M.E., J.K.M., S.A.A., J.S., F.P., and P.C. Software: G.M.-E. and M.G. Supervision: P.D., S.A., and P.C. Validation: G.M.-E., P.D., N.Z., S.A., and P.C. Visualization: G.M.-E. Writing—original draft: G.M.-E., P.D., N.Z., S.A., and P.C. Writing—review and editing: G.M.-E., P.D., N.Z., U.S., M.B., P.M.B., Z.M., M.G., N.A., E.J.A., B.d.G., J.B., U.P.-B., A.A.V., R.J.M., E.R., S.H., M.E., M.C., J.K.M., S.A.A., J.S., F.P., and P.C.

Author Disclosure Statement

G.M.E.'s position at King's College London is funded by a research grant from Novo Nordisk as part of their contribution to the Hypo-RESOLVE consortium. S.A.A. has served on advisory boards for Novo Nordisk and Medtronic and has spoken at an educational symposium sponsored by Sanofi. M.L.E. has served on advisory boards and/or received lecture fees and/or research support from Novo Nordisk, Eli Lilly, AstraZeneca, Medtronic, Dexcom, Ypsomed, Abbott Diabetes Care, Roche, NGM Pharma, Zucara, and Pila Pharma. U.P.B. has served on advisory boards and has received lecture fees from Sanofi and Novo Nordisk.

J.K.M. is a member of RM and has served on advisory boards of Abbott Diabetes Care, Becton-Dickinson, Boehringer Ingelheim, Eli Lilly, Embecta, Medtronic, Novo Nordisk A/S, Roche Diabetes Care, Sanofi-Aventis, and Viatris and has received speaker honoraria lecture fees from A. Menarini Diagnostics, Abbott Diabetes Care, AstraZeneca, Boehringer Ingelheim, Dexcom, Eli Lilly, MedTrust, MSD, Novo Nordisk A/S, Roche Diabetes Care, Sanofi, Servier, Sanofi, and Ypsomed. She is a shareholder of decide Clinical Software GmbH and elyte Diagnostics where she also serves as CMO. B.D.G. has received research support from Novo Nordisk. E.R. has served as consultant/advisor for Abbott, Air Liquide SI, AstraZeneca, Boehringer-Ingelheim, Dexcom, Eli-Lilly, Hillo, Insulet, Medirio, Novo Nordisk, Roche, Sanofi-Aventis, Tandem, and received research support from Dexcom and Tandem.

J.S. has served on advisory boards for Janssen, Medtronic, Roche Diabetes Care, and Sanofi Diabetes; her research group (Australian Centre for Behavioural Research in Diabetes [ACBRD]) has received honoraria for this advisory board participation and has also received unrestricted educational grants and in-kind support from Abbott Diabetes Care, AstraZeneca, Medtronic, Roche Diabetes Care, and Sanofi Diabetes. J.S. has also received sponsorship to attend educational meetings from Medtronic, Roche Diabetes Care, and Sanofi Diabetes, and consultancy income or speaker fees from Abbott Diabetes Care, AstraZeneca, Medtronic, Novo Nordisk, Roche Diabetes Care, and Sanofi Diabetes. P.C. has received personal fees Abbott Diabetes Care, Insulet, Dexcom, Novo Nordisk, AstraZeneca, Medtronic, Roche Diabetes Care, and Sanofi Diabetes. Research funding support from Abbott Diabetes Care, Medtronic, and Novo Nordisk.

Funding Information

This study represents independent research supported by the National Institute for Health and Care Research (NIHR) King's Clinical Research Facility and the NIHR Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King's College London. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR, or the Department of Health and Social Care. Hypo-RESOLVE has received funding from the Innovative Medicines Initiative 2 Joint Undertaking (JU) under grant agreement No 777460. The JU receives support from the European Union's Horizon 2020 research and innovation programme and EFPIA and T1D Exchange, JDRF, International Diabetes Federation (IDF), and The Leona M. and Harry B. Helmsley Charitable Trust. The industry partners supporting the JU include Abbott Diabetes Care, Eli Lilly, Medtronic, Novo Nordisk, and Sanofi-Aventis.

The funder had no role in the design of the project or its WPs, the collection or analysis of data, the writing of the article, or the decision to submit for publication.

References

Graveling

, Frier

. The risks of nocturnal hypoglycaemia in insulin-treated diabetes. Diabetes Res Clin Pract, 2017; 133:30–39; doi: 10.1016/j.diabres.2017.08.012

Jones

, Porter

, Sherwin

, et al. Decreased epinephrine responses to hypoglycemia during sleep. N Engl J Med, 1998; 338(23):1657–1662; doi: 10.1056/NEJM199806043382303

Banarer

, Cryer

. Reduced awakening from sleep during hypoglycemia. Diabetes, 2003; 30(10):1195–1203.

Jauch-Chara

, Schultes

. Sleep and the response to hypoglycaemia. Best Pract Res Clin Endocrinol Metab, 2010; 24(5):801–815; doi: 10.1016/j.beem.2010.07.006

Edelman

, Blose

. The impact of nocturnal hypoglycemia on clinical and cost-related issues in patients with type 1 and type 2 diabetes. Diabetes Educ, 2014; 40(3):269–279.

Chatwin

, Broadley

, Valdersdorf Jensen

, et al. Never again will i be carefree’: A qualitative study of the impact of hypoglycemia on quality of life among adults with type 1 diabetes. BMJ Open Diabetes Res Care, 2021; 9(1):1–10.

Brown

, Reichert

, Valliere

, Webster-Bogaert

, et al. Living with hypoglycemia: An exploration of patients' emotions: Qualitative findings from the InHypo-DM Study, Canada. Diabetes Spectr, 2019; 32(3):270–276.

Khunti

, Alsifri

, Aronson

, et al. Rates and predictors of hypoglycaemia in 27 585 people from 24 countries with insulin-treated type 1 and type 2 diabetes: The global HAT study. Diabetes Obes Metab, 2016; 18(9):907–915; doi: 10.1111/dom.12689

Battelino

, Danne

, Edelman

, et al. Continuous glucose monitoring-based time-in-range using insulin glargine 300 units/ml versus insulin degludec 100 units/ml in type 1 diabetes: The head-to-head randomized controlled InRange trial. Diabetes Obes Metab, 2023; 25(2):545–555.

10.

Brosen

JMB

, Agesen

, Alibegovic

, et al. Continuous glucose monitoring-recorded hypoglycemia with insulin degludec or insulin Glargine U100 in people with type 1 diabetes prone to nocturnal severe hypoglycemia. Diabetes Technol Ther, 2022; 24(9):643–654.

11.

Larcher

, Benhamou

P-Y

, Pépin

J-L

, Borel

A-L

. Sleep habits and diabetes. Diabetes Metab, 2015; 41(4):263–271; doi: 10.1016/j.diabet.2014.12.004

12.

Quinn

, Hadjiconstantinou

, Brady

, et al. Chronotype and well-being in adults with established type 2 diabetes: A cross-sectional study. Diabet Med, 2022; 39(3):e14690; doi: 10.1111/dme.14690

13.

Robbins

, Seixas

, Walton Masters

, et al. Sleep tracking: A systematic review of the research using commercially available technology. Curr Sleep Med Reports, 2019; 5(3):156–163; doi: 10.1007/s40675-019-00150-1

14.

Haghayegh

, Khoshnevis

, Smolensky

, et al. Accuracy of wristband fitbit models in assessing sleep: Systematic review and meta-analysis. J Med Internet Res, 2019; 21(11):e16273.

15.

Dong

, Yang

, Guo

, et al. Validation of Fitbit Charge 4 for assessing sleep in Chinese patients with chronic insomnia: A comparison against polysomnography and actigraphy. PLoS One, 2022; 17:1–14; doi: 10.1371/journal.pone.0275287

16.

Renerts

, Doheny

, Baumann

, et al. F63 Validation of fitbit charge 4 for sleep monitoring in participants with Huntington's disease. In: F: Clinical studies: case reports, observational studies and trials. BMJ Publishing Group Ltd;, 2022; pp. A59.1–A59; doi: 10.1136/jnnp-2022-ehdn.154

17.

de Zambotti

, Goldstone

, Claudatos

, et al. A validation study of Fitbit Charge 2^TM compared with polysomnography in adults. Chronobiol Int, 2018; 35(4):465–476; doi: 10.1080/07420528.2017.1413578

18.

Flatt

, Peleckis

, Dalton-Bakes

, et al. Automated insulin delivery for hypoglycemia avoidance and glucose counterregulation in long-standing type 1 diabetes with hypoglycemia unawareness. Diabetes Technol Ther, 2023; 25(5):302–314.

19.

Malone

, Peleckis

, Grunin

, et al. Characterizing glycemic control and sleep in adults with long-standing type 1 diabetes and hypoglycemia unawareness initiating hybrid closed loop insulin delivery. J Diabetes Res, 2021; 2021:1–8.

20.

de Galan

, McCrimmon

, Ibberson

, et al. Reducing the burden of hypoglycaemia in people with diabetes through increased understanding: design of the Hypoglycaemia REdefining SOLutions for better liVEs (Hypo-RESOLVE) project. Diabet Med, 2020; 37(6):1066–1073.

21.

Divilly

, Zaremba

, Mahmoudi

, et al. Hypo-METRICS: Hypoglycaemia—MEasurement, ThResholds and ImpaCtS—A multi-country clinical study to define the optimal threshold and duration of sensor-detected hypoglycaemia that impact the experience of hypoglycaemia, quality of life and health economic o. Diabet Med, 2022; 39(9):1–11.

22.

Søholm

, Broadley

, Zaremba

, et al. Investigating the day-to-day impact of hypoglycaemia in adults with type 1 or type 2 diabetes: Design and validation protocol of the Hypo-METRICS application. BMJ Open, 2022; 12(2):1–12.

23.

Danne

, Nimri

, Battelino

, et al. International consensus on use of continuous glucose monitoring. Diabetes Care, 2017; 40(12):1631–1640.

24.

Feupe

, Frias

, Mednick

, et al. 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.

25.

Botella-Serrano

, Velasco

, Sánchez-Sánchez

, et al. Evaluating the influence of sleep quality and quantity on glycemic control in adults with type 1 diabetes. Front Endocrinol (Lausanne), 2023; 14:1–14.

26.

Chontong

, Saetung

, Reutrakul

. Higher sleep variability is associated with poorer glycaemic control in patients with type 1 diabetes. J Sleep Res, 2016; 25(4):438–444.

27.

Henson

, Rowlands

, Baldry

, et al. Physical behaviors and chronotype in people with type 2 diabetes. BMJ Open Diabetes Res Care, 2020; 8(1):e001375; doi: 10.1136/bmjdrc-2020-001375

28.

Valiensi

, Folgueira

, Diez

, et al. Is being a lark healthier for patients with type 1 diabetes mellitus?. Sleep Sci, 2023; 16(1):75–83; doi: 10.1055/s-0043-1767749

29.

Zou

, Zhou

, Yan

, et al. Chronotype, circadian rhythm, and psychiatric disorders: Recent evidence and potential mechanisms. Front Neurosci, 2022; 16:1–24.

30.

Rowlands

, Henson

, Coull

, et al. The impact of COVID-19 restrictions on accelerometer-assessed physical activity and sleep in individuals with type 2 diabetes. Diabet Med, 2021; 38(10):e14549; doi: 10.1111/dme.14549

31.

The DCCT Research Group. Epidemiology of severe hypoglycemia in the diabetes control and complications trial. Am J Med, 1991; 90(1):450–459.

32.

Pedersen-Bjergaard

, Alsifri

, Aronson

, et al. Comparison of the HAT study, the largest global hypoglycaemia study to date, with similar large real-world studies. Diabetes Obes Metab, 2019; 21(4):844–853.

33.

Zhu

, Kapella

, Zhao

, Fritschi

. Intra-individual variability in sleep is related to glycaemic control in adults with type 2 diabetes. J Adv Nurs, 2020; 76(4):991–998; doi: 10.1111/jan.14290

34.

Zhu

, Bronas

, Fritschi

. Sleep assessment in aging adults with type 2 diabetes: Agreement between actigraphy and sleep diaries. Sleep Med, 2018; 46:88–94; doi: 10.1016/j.sleep.2018.03.008

35.

Maynard

, Appleman

, Cronin-Golomb

, Neargarder

. Objective measurement of sleep by smartphone application: Comparison with actigraphy and relation to self-reported sleep. Explor Med, 2021; 2(5):382–391.