Impaired Awareness and Symptomatology of Hypoglycemia in Adults With Type 1 Diabetes Using Continuous Glucose Monitoring

Abstract

Context and Objective

Sparse large-scale studies have characterized hypoglycemia symptomatology in adults with type 1 diabetes (T1D) who use continuous glucose monitoring (CGM). This research aimed to evaluate the relationship of impaired awareness of hypoglycemia (IAH) with hypoglycemia symptomatology and frequency in this population.

Design

A cross-sectional survey was conducted in 2023. HypoA-Q was used to evaluate hypoglycemia frequency, symptomatology, and awareness.

Setting & Participants

Registrants who used CGM in the T1D Exchange, a U.S. national T1D registry.

Results

Surveys were completed by 1480 adults with T1D who used CGM (53% female; mean diabetes duration 26 years). Compared to those with intact hypoglycemia awareness, IAH was associated with less frequent hypoglycemia symptom presentation across various glucose levels and lower glucose concentrations for first presenting hypoglycemia symptoms when awake (P < .05 for all). More than 70% of individuals with IAH did not regularly experience symptoms during episodes with glucose <54 mg/dL. When asleep, those with IAH were less frequently awoken by symptoms, more frequently awoken by others who recognized their hypoglycemia, and more frequently acknowledged hypoglycemia after an episode (P < .05 for all) despite using CGM. With reduced symptoms, both when awake and asleep, those with IAH experienced more hypoglycemic episodes that they were unable to self-treat than those with intact awareness (P < .05).

Conclusion

IAH continues to be associated with a significant reduction in detection of hypoglycemia symptoms, both when awake and asleep, in adults using CGM. Current diabetes technologies do not fully protect adults with T1D from hypoglycemia.

Hypoglycemia is an acute, common, and potentially lethal complication of exogenous insulin in people with type 1 diabetes (1). Impaired awareness of hypoglycemia (IAH), a condition characterized by the reduced ability to perceive hypoglycemia symptoms, raises the risk of developing severe hypoglycemia (2). Hypoglycemia symptomatology in people living with type 1 diabetes has been extensively investigated, including the establishment of the physiological glucose thresholds of symptom responses (3), the prevalence of IAH (4), and the difference in the glucose thresholds of symptom responses between those with intact and impaired awareness (5). However, these studies were mostly performed prior to the widespread use of continuous glucose monitoring (CGM). In addition, many of these studies utilized hyperinsulinemic-hypoglycemic clamps (6), the gold standard for assessing hypoglycemia symptoms in controlled settings. To investigate hypoglycemia in real-world settings, research has typically involved requesting regular self-monitoring of blood glucose levels via finger-prick glucose meters and self-completion of postevent hypoglycemia symptom diaries or logs (4, 7). Recent research has involved blinded CGM and symptom experiences collected daily via a smartphone App (8).

Despite using advanced diabetes technologies, such as CGM and automated insulin delivery (AID; ie, insulin delivery through insulin pumps that can be modulated based on sensor glucose data (9)) devices, clinically significant hypoglycemia and IAH continue to affect from one quarter to one third of adults with type 1 diabetes (10-12). People with IAH may manage asymptomatic hypoglycemic episodes suboptimally due to not experiencing symptoms as confirmation of CGM glucose information or as a somatic prompt to treat hypoglycemia (13). Behavioral interventions have the potential to improve hypoglycemia awareness and prevent severe events, including among individuals using CGM (14). Thus, a clearer understanding of the symptoms associated with hypoglycemia is crucial for individuals who use CGM. Only a few studies, including a hypoglycemic clamp study showing partial recovery of hypoglycemia awareness with CGM (15) as well as a recent innovative ecological momentary assessment (16-18) project, have evaluated hypoglycemia symptomatology in CGM users. Sample size, however, was limited in these studies due to the nature of the assessment methods. A large-scale survey may further illuminate the experiences of hypoglycemia symptomatology in people living with type 1 diabetes and using CGM.

The aim of this study was to use an instrument validated for assessing hypoglycemia experiences (ie, HypoA-Q) to characterize hypoglycemic events and symptoms, and examine their associations with IAH in people with type 1 diabetes who use CGM.

Materials and Methods

Design and Participants

A cross-sectional survey was conducted between July and December 2023. This study was approved by the University of Michigan institutional review board (HUM#00232351). All participants provided informed consent prior to completing study activities. The study is reported in accordance with the STROBE guidelines (19).

Participants were recruited via email, including brief introductory information, through T1D Exchange, a national type 1 diabetes patient registry in the United States (20). Adults living with type 1 diabetes were eligible to complete the survey. Interested registrants could choose to open the survey website, review the study information, provide consent, and complete the survey. Only participants reporting ongoing use of CGM were eligible for the analysis.

Data Collection

A survey webpage built with REDCap (21) was used to administer the survey and collect responses. Participants’ self-reported demographic information (ie, age, gender, race, ethnicity, and education level) and diabetes information (ie, diabetes duration, self-reported glycated hemoglobin), and CGM and AID use statuses) were collected.

Participants’ hypoglycemia experiences and symptoms were assessed using the Hypoglycemia Awareness Questionnaire (HypoA-Q) (22). The HypoA-Q is a profile measure, including 20 items covering the various hypoglycemia experiences when awake and asleep, their frequency, and the levels at which symptoms occur. In addition, 5 items assessing perceived awareness of hypoglycemia symptoms are summed to form an “Impaired Awareness” scale (22, 23); scores of ≥12 have been shown to indicate IAH (24). The “When awake” section also includes 1 item (item #5, which consists of 4 subitems—5a, 5b, 5c, and 5d) assessing the glucose levels at which types of hypoglycemia symptoms are experienced (“Symptom Level” scale) and 1 item (item #6, which consists of 3 subitems—6a, 6b, and 6c) assessing how frequently hypoglycemia symptoms have been experienced at various glucose levels (“Symptom Frequency” scale). The “When asleep” section also assesses hypoglycemia management experiences and mechanisms for recognizing hypoglycemia that occurs during sleep (ie, awoken by hypoglycemia symptoms, by someone else recognizing hypoglycemia, and by diabetes technology), as well as major problems due to hypoglycemia (eg, seizure, tongue biting, fall, collapse, incontinence).

The “When awake” Symptom Frequency and Symptom Level scale scores (ie, item Q5-6) were assigned and calculated based on the updated scale scoring system (25). All other scores were assigned according to the original HypoA-Q scoring system (22).

Statistical Analysis

A sample of 377 respondents was determined to adequately represent the pool of 18 000 adult registrants in the T1D Exchange Registry at a 95% confidence level and a margin of error of 5%. We opted to oversample to enhance estimates’ precision and to account for non-CGM users and potential incomplete or duplicate responses. Illogical values were removed on data review and reported as missing data; all valid responses were retained for analysis. Descriptive analyses were conducted to summarize the study results as means and standard deviations unless otherwise noted. Univariate and multivariate linear regressions were performed to evaluate the relationships between IAH status and the following attributes, without and with adjusting for AID use status: hypoglycemia frequency; frequency of needing assistance for treatment; symptomatic glucose levels and frequency when awake; HypoA-Q Symptom Frequency and Symptom Level scale scores; mechanisms for acknowledging hypoglycemia that occurs when asleep; and problems secondary to hypoglycemia that occurs when asleep. All missing data were considered missing completely at random. No missing data were identified in eligible participants’ hypoglycemia experience reporting. P < .05 was considered as statistically significant (2-tailed).

Results

Demographics

After removing duplicate and incomplete responses and individuals not using CGM, responses from N = 1480 eligible participants remained for analysis (Supplementary Fig. 1) (26). The participants’ mean ± SD age was 46 ± 15 years, diabetes duration was 26 ± 16 years, and glycated hemoglobin was 6.7 ± 1.1% (50 ± 12 mmol/mol); 53% were women, 95% were White, and 94% were non-Hispanic (Table 1). About 95% of participants had been using CGM for at least 1 year and 72% were using AID. Of all participants, 230 (16%) were found to have IAH.

Hypoglycemia Frequency, Severity and Symptom Experiences When Awake

Participants with intact awareness of hypoglycemia (AH) and IAH reported a comparable frequency of hypoglycemic episodes when awake in the past 6 months (Table 2). However, participants with IAH reported more episodes where they were unable to self-treat and requiring someone to administer food treatment or inject glucagon (P < .05 for all). Notably, participants with IAH had approximately double the prevalence of experiencing at least 1 episode that they were unable to self-treat (IAH 37% vs AH 17%), in addition to needing for assistance in administering food treatment (IAH 31% vs AH 16%) or glucagon (IAH 8.3% vs AH 3.4%) in the past 6 months.

Compared to participants with AH, those with IAH reported fewer symptomatic episodes across all hypoglycemic glucose levels (P < .05 for all; Table 3). This finding was largely due to fewer individuals with IAH reporting “always” experiencing symptoms but more frequently “sometimes,” “rarely,” or “never” being aware of symptoms at each level of biochemical hypoglycemia. When glucose levels were between 45 and 53 mg/dL and below 45 mg/dL, 55% and 61% of those with AH always detected symptoms, compared to only 11% and 15% of the IAH group, respectively. Moreover, participants with IAH developed not only autonomic, but also neuroglycopenic, and other, symptoms at lower glucose levels than participants with AH (P < .05 for all; Supplementary Table 1) (26), evidencing a downward shift in the glucose levels necessary for symptoms (Fig. 1). Statistical significance of all results was unaffected by adjusting for AID use status.

IAH was associated with higher Symptom Frequency scale scores before (estimate 95% CI, 4.874-6.047) and after (estimate 95% CI, 4.876-6.051) adjusting for AID use status. Similarly, IAH status was associated with higher Symptom Level scale scores before (estimate 95% CI, 2.118-2.929) and after (estimate 95% CI, 2.106-2.917) adjusting for AID use status.

Hypoglycemia Frequency and Symptom Experiences When Asleep

Participants with IAH reported overall more hypoglycemic episodes when asleep, and more episodes unable to self-treat on waking, and requiring assistance to administer food treatment or glucagon (P < .05 for all; Table 4). Again, participants with IAH had approximately double the prevalence of experiencing at least 1 hypoglycemic episode that they were unable to self-treat on waking (IAH 33% vs AH 18%), in addition to need for assistance in administering food treatment (IAH 21% vs AH 9.8%) or glucagon (IAH 6.5% vs AH 2.5%). About 12% of those with IAH also reported having experienced a hypoglycemic episode leading to major problems, such as seizure, fall, and incontinence (vs 3% of those with AH).

Participants with IAH reported being awoken by hypoglycemia symptoms less frequently than those with AH but reported more frequently being awoken by others who recognized hypoglycemia (P < .05 for both; Fig. 2). They also more frequently recognized hypoglycemia only after an episode (P < .05; Supplementary Table 2) (26). No relationship was observed between IAH and the frequency of being awoken by advanced diabetes technologies. Approximately 7.9% of participants with AH and 8% with IAH reported that they were never or rarely awoken by these technologies during hypoglycemia. Statistical significance of all results remained unaffected by adjusting for AID use status.

Discussion

In this national hypoglycemia symptomatology study of adults with type 1 diabetes using CGM, we observed that IAH affected the frequency of symptom presentation and glycemic thresholds for developing autonomic, neuroglycopenic, and other symptoms while awake. The majority of individuals with IAH did not regularly experience symptoms, even with hypoglycemia in the dangerous, neuroglycopenic range (ie, <54 mg/dL). When asleep, IAH was associated with being awoken by symptoms less frequently and recognizing hypoglycemia more often after the episode. Hypoglycemia being recognized by someone else differentiated those with IAH from those with intact awareness, despite all participants using CGM. Given these differences, over the past 6 months, 38% of individuals with IAH had experienced hypoglycemic episodes that they were unable to self-treat while awake. Similarly, 33% of participants with IAH had experienced episodes during sleep that they were unable to self-treat upon waking. IAH also correlated with more frequent episodes requiring assistance for treatment and glucagon administration, both when awake or asleep, and with episodes that led to major problems when asleep.

The observed relationships between IAH and hypoglycemic episodes requiring assistance are largely consistent with recent research demonstrating the ongoing issue of IAH despite using CGM (11, 12, 23). Additionally, we identified that people with IAH had no more overall hypoglycemic episodes but experienced more serious episodes when awake than those with intact awareness, aligning with real-world CGM data (27). Thus, although CGM may help with timely detection and prevention of some of these milder hypoglycemic episodes, people with IAH continues to face higher risks for developing more serious events, potentially from their reduced counterregulatory hormone responses (28) and suboptimal self-management of asymptomatic episodes (13). It is also particularly worrisome that about one third of individuals with IAH, despite using CGM, experienced hypoglycemia while asleep and were unable to self-treat, consistent with the previously identified relationship between IAH and more serious nocturnal hypoglycemia in CGM users (27). It remains to be determined how changes in the setting of glucose threshold for hypoglycemia alarms (eg, 80 mg/dL) during sleep and whether using predictive hypoglycemia alarms that can wake people with IAH or their carers early in such events (29) would impact hypoglycemia reporting. Psychoeducation or behavioral training to improve awareness and promote self-management strategies to reduce severe hypoglycemia risk (30-32), as well as AID to enable partial recovery of glucose counterregulatory mechanisms (33), may also serve as potential rescue strategies.

In our study, we observed that among the group categorized as having intact awareness based on HypoA-Q IA with the hypoglycemic clamp–determined score cut-point (ie, Impaired Awareness scale score <12) (24), some still rarely or never experienced symptoms even at glucose levels <54 mg/dL. Our study supports the assertion that IAH exists on a continuum and may be dynamic (10, 34, 35), and detailed characterization of hypoglycemia awareness for each individual may be necessary. In addition, alternate hypoglycemia awareness measures (eg, HypoA-Q IA, Gold (36), Clarke (37), and Pedersen-Bjergaard instruments (38)) do not effectively discriminate between episodes experienced while awake and while sleeping. Our research demonstrates the utility of HypoA-Q for systematic collection of details about the frequency, severity, symptoms, and awareness of hypoglycemia that occurs while awake and asleep. During our study conduct, we observed that the HypoA-Q administration was neither time- nor labor-intensive. This instrument may hence be suitable for use in large cohorts, now including CGM users, in parallel with or replacing other hypoglycemia assessment methods.

In our analysis, adjusting for AID use did not significantly alter the results. Our findings contribute to the enduring controversy regarding whether AID provides additional benefits in hypoglycemia reduction and hypoglycemia awareness improvement compared with standalone CGM or other technologies (39-43), though the current data are limited by responder bias. Preliminary work prospectively investigating AID implementation in individuals with IAH demonstrates the potential for achieving further hypoglycemia reduction and associated restoration of autonomic symptom responses to insulin-induced hypoglycemia (33), while data from large-scale trial research are pending.

To the best of our knowledge, this study represents the first large-scale research of hypoglycemia symptomatology in CGM users. The strengths of this study include its nationwide sampling, large sample size, and minimal missing data. In addition, most participants had longitudinal CGM use history, and administering HypoA-Q enabled a comprehensive hypoglycemia symptomatology and experience assessment.

Our study has several limitations. Recall bias may affect the accuracy of self-reported data. Because participants were exclusively recruited from a patient registry and were all CGM users, the generalizability of the finding to a wider type 1 diabetes population may be constrained. Additionally, the underrepresentation of minority groups prevented assessment of applicability of our findings across diverse demographics. Finally, the completion of a survey involving multiple instruments, including demographics, severe hypoglycemia history, and the 20-item HypoA-Q questionnaire, may result in a sample skewed toward more dedicated participants. The high survey completion rate, indicating the survey's feasibility, may yet offset some of this concern.

In conclusion, our study suggests that, despite use of advanced diabetes technologies, hypoglycemia and IAH continue to be challenging and to negatively affect the lives of people with type 1 diabetes. Technological awareness is useful but does not ensure full safety among those with IAH, neither when awake nor asleep. Additional interventions in addition to the use of current technologies are thus needed to minimize serious hypoglycemic episodes in people with IAH. Research documenting real-life experiences with hypoglycemia using tools such as the HypoA-Q can provide descriptive parallel appraisals complementary to hypoglycemic clamp–based symptomatology studies.

Acknowledgments

We greatly appreciate support from the T1D Exchange and all the study participants for completing this study.

Funding

This work was supported by the National Institute of Diabetes and Digestive and Kidney Diseases (K23DK129724) and Michigan Center for Clinical and Translational Research (P30DK092926). M.R.R. is supported in part by U01 DK135120. REDCap was supported by NCATS UL1TR00240.

Author Contributions

Y.K.L., M.R.R., S.A.A., J.S., and J.A.M.S. were involved in study conception, design, and interpretation of results. Y.K.L. and E.H. collected the data. Y.K.L. and W.Y. conducted the analysis. Y.K.L. wrote the first draft of the manuscript and is the guarantor of this work. All authors edited, reviewed, and approved the final version of the manuscript.

Disclosures

J.S. owns the copyright for HypoA-Q. All other authors have no conflicts of interest relevant to this study.

Data Availability

Data are available on request to Yu Kuei Lin ([email protected]).

Prior Presentation

The study has not been presented in prior studies.

References

Abbreviations

 
• AH

  awareness of hypoglycemia

 
• AID

  automated insulin delivery

 
• CGM

  continuous glucose monitoring

 
• IAH

  impaired awareness of hypoglycemia