Use of Low-Glucose Alerts and Their Impact on Continuous Glucose Monitoring Outcomes in Real-World Use of the MiniMed 780G System

Abstract

Objective:

Hypoglycemia remains a major concern in insulin therapy and has limited intensive glycemic control. Although automated insulin delivery substantially reduces hypoglycemia risk, relatively high low-glucose alert (LA) thresholds (LGAT) remain commonly used. This study evaluated the impact of user-set LGAT on glycemic outcomes.

Research Design and Methods:

This retrospective cohort study analyzed anonymized real-world data from 169,487 MiniMed 780G users with type 1 diabetes across Europe, the Middle East, and Africa (August 1, 2024–July 31, 2025). Use of LAs, user-set thresholds, and glycemic metrics was evaluated.

Results:

Most users (96%) enabled LGAT, and 36% enabled predictive LGAT. Median, 25th and 75th percentiles of time below 54 and 70 mg/dL remained low and within recommended targets across all alert thresholds overall, in users aged ≤15 years, and overnight. Hypoglycemia targets were achieved by >82% and >90% of users with thresholds <60 and ≥60 mg/dL, respectively. In contrast, time in range (TIR) 70–180 mg/dL and hyperglycemia targets were more frequently achieved with LGAT <60 mg/dL, with 70% and 61% meeting these targets, respectively. Overall, 43% of users with LGAT <60 mg/dL met all glycemic targets versus 31% with thresholds >80 mg/dL.

Conclusions:

A LGAT below 60 mg/dL is generally recommended, as it is associated with increased TIR, reduced hyperglycemia, and low time in hypoglycemia. A threshold between 60 and 70 mg/dL may provide a balanced option for those with greater hypoglycemia concerns. These recommendations can optimize system usability and glycemic outcomes while maintaining safety and minimizing alarm burden.

Keywords

automated insulin delivery real-world evidence closed-loop system low-glucose alert thresholds hypoglycemia glucose management indicator advanced hybrid closed-loop type 1 diabetes

Article Highlights

•

Why did we undertake this study?

Hypoglycemia remains a major concern in insulin therapy and has limited intensive glycemic control. Low-glucose alert thresholds (LGAT) are configured to provide warnings when glucose levels fall below a threshold.

•

What is the specific question(s) we wanted to answer?

What is the impact of LGAT in glycemic control with automated insulin delivery (AID) that reduces hypoglycemia risk?

•

What did we find?

LGAT below 60 mg/dL are generally recommended as they are associated with low time in hypoglycemia, reduced hyperglycemia, and increased time in range. A threshold between 60 and 70 mg/dL may provide a balanced option for those with greater hypoglycemia concerns.

•

What are the implications of our findings?

These recommendations can optimize AID usability and glycemic outcomes while maintaining safety and minimizing alarm burden.

Introduction

Hypoglycemia, and particularly nocturnal hypoglycemia, has traditionally been considered the most feared adverse effect of insulin therapy and a major limiting factor in achieving optimal glycemic control in diabetes.¹ Beyond its immediate physiological consequences, hypoglycemia is often described by patients as a profoundly distressing experience. Data from the JDRF Continuous Glucose Monitoring (CGM) trial demonstrated that the mean duration of nocturnal hypoglycemia is approximately 81 min.² In addition, reports of the “dead in bed” syndrome, estimated to account for 3%–6% of deaths among individuals with diabetes younger than 40 years,³ have further reinforced concerns regarding nocturnal hypoglycemia. In the United States, this corresponds to an estimated 35–130 deaths annually, contributing to a pervasive fear of hypoglycemia among both people living with diabetes and healthcare providers.

Early strategies to mitigate hypoglycemia risk focused on the use of CGM alerts configured to provide early warnings when glucose levels fell below a preset threshold or were predicted to do so. Although such alerts were associated with reductions in hypoglycemia, they did not eliminate it.⁴ Moreover, frequent alerts introduced additional user burden and alarm fatigue, potentially diminishing their effectiveness in preventing hypoglycemia.⁵

Evidence from the Diabetes Control and Complications Trial showing that severe hypoglycemia occurred more than threefold more frequently in intensively treated participants than in conventionally treated individuals contributed to the long-standing belief that severe hypoglycemia could be mitigated primarily by relaxing glycemic targets and accepting higher HbA1c levels.¹ However, recent advances in automated insulin delivery (AID) systems have substantially challenged this paradigm.

The introduction of features such as automatic basal insulin suspension, at or before predicted low glucose levels, has significantly reduced time spent in hypoglycemia as well as the incidence of severe hypoglycemic events.⁶ More recently, the introduction of the MiniMed 780G AID system (Medtronic, Northridge, CA), which incorporates automated correction boluses and hypoglycemia prevention algorithms based on adaptive meal and correction models,⁷ has enabled improved glycemic outcomes while simultaneously reducing hypoglycemia risk. This technology effectively decouples improvements in mean glucose and glucose management indicator (GMI) from increased hypoglycemia risk,^8,9 enabling safe and effective therapy across large real-world populations.¹⁰ Despite these technological advances, clinical practice and educational frameworks have not always evolved at the same pace.

One example is the continued use of relatively high CGM low-glucose alarm thresholds, a practice largely inherited from earlier therapeutic paradigms predating advanced AID systems such as the MiniMed 780G. Current recommendations, reinforced through training materials, educational programs, and some clinical guidelines, often differ between pediatric and adult populations.¹¹ Lin et al. demonstrated that among individuals with type 1 diabetes treated with either multiple daily injections or open-loop insulin pump therapy, higher CGM low-glucose alert (LA) thresholds (LGAT) were associated with reduced time spent in hypoglycemia but at the cost of increased time in hyperglycemia, with a recommended cutoff of approximately 73 mg/dL.¹² Although individualized threshold settings have long been advocated,¹³ thresholds of 70 mg/dL or higher remain commonly used in clinical practice.^14,15

This study was designed to provide high-quality evidence regarding the use of CGM LGAT and their impact on individuals using the MiniMed 780G AID system within a large cohort of real-world users. The objective was to generate data-driven recommendations that optimize both safety and usability of the system, reduce unnecessary alarm burden, and support improved glycemic outcomes across diverse patient populations spanning different ages and geographic regions.

Research Design and Methods

Data source and study population

This retrospective observational cohort study utilized anonymized real-world data from users of the MiniMed 780G system uploaded to CareLink™ Personal between August 1, 2024, and July 31, 2025. The dataset included individuals living with type 1 diabetes across Europe, the Middle East, and Africa (EMEA).

To be eligible for inclusion, individuals were required to have a registered CareLink Personal account within the EMEA region and to have provided consent for their anonymized data to be used for research purposes. Users were additionally required to have at least 10 days of sensor glucose (SG) data available following the first activation of the advanced hybrid closed-loop (AHCL) system. All available CGM data from eligible users were included in the analysis, regardless of whether the system was operating in AHCL mode or open-loop mode during the observation period.

CGM alert settings

The MiniMed 780G system allows users to configure two types of LAs: the LA and the predictive LA (PLA).

The LA is triggered when the current SG value reaches or falls below a user-defined threshold, notifying the user that glucose is already low and requires attention. This alert is based solely on the measured SG value, and the threshold can be programmed within the 50–90 mg/dL range permitted by the system.

The PLA is triggered when the system algorithm estimates that SG is projected to reach the low threshold within a predefined time window of 30 min, even if the current glucose value remains above the threshold. This alert relies on trend analysis and the rate of change of SG and therefore provides an early warning before hypoglycemia occurs. When the system operates in SmartGuard™ mode, the algorithm may automatically reduce or suspend insulin delivery in response to predicted hypoglycemia in addition to issuing the alert.

Statistical analyses

We evaluated the frequency of use of LA and PLA, the distribution of user-defined alert thresholds, and glycemic outcomes including time in range (TIR; 70–180 mg/dL), time in tight range (TITR; 70–140 mg/dL), time above 180 mg/dL (TA180), time above 250 mg/dL (TA250), time below 70 mg/dL (TB70), and time below 54 mg/dL (TB54). These outcomes were analyzed according to the type of glucose alert setting in the overall cohort, among users aged ≤15 years, and during nighttime, defined as the period between 00:00 and 06:00.

The impact of user-set alert thresholds on glycemic outcomes was assessed by comparing median values of TIR metrics across alert threshold categories of 54–60, 60–65, 65–70, 70–75, 75–80, 80–85, and 85–90 mg/dL. Differences across groups were evaluated using the Kruskal–Wallis rank-sum test, and the Benjamini–Hochberg procedure was applied to adjust for multiple comparisons.

We further identified alert thresholds associated with achieving recommended clinical glycemic targets, defined as TIR >70%, TA180 <25%, TA250 <5%, TB70 <4%, and TB54 <1%. Receiver operating characteristic (ROC) analysis using the Youden Index was used to determine the cutoff value of the user-set alert threshold that maximized the true positive rate while minimizing the false positive rate for meeting all clinical targets simultaneously. Spearman correlation between the alert thresholds and time in hypoglycemia was performed.

All statistical analyses were conducted using R software (version 4.3.1). A two-sided P value <0.05 was considered statistically significant.

Results

A total of 169,487 individuals met the inclusion criteria and constituted the overall cohort, of whom 39,597 were aged ≤15 years (younger cohort).

Use of glucose alerts in the overall cohort

Among all users, 163,059 (96%) used LA and 61,350 (36%) used PLA. A total of 60,572 (36%) used both LA and PLA, 102,487 (60%) used only LA, 778 (0.5%) used only PLA, and 5650 (3.3%) did not use any alert.

Among LA users, the median [25th–75th percentile] percentage of time with LA enabled was 100% [100%–100%], while among PLA users, the median time with PLA enabled was 100% [69%–100%]. During the observation period, LA and PLA users set equivalent median alert thresholds, 70 mg/dL [68.5–72 mg/dL] for LA, and 70 mg/dL [67–74 mg/dL] for PLA.

Glycemic outcomes were similar between LA and PLA users. Median values for TB54, TB70, TITR, TIR, TA180, TA250, and GMI were 0.2%, 1.4%, 47.3%, 72.8%, 25.4%, 4.8%, and 6.9%, respectively, in LA users and 0.2%, 1.3%, 46.9%, 72.7%, 25.5%, 4.8%, and 7%, respectively, in PLA users (Table 1).

Table 1.

Glucose Alert Use, the Time in Different Ranges in Low-Glucose Alert, and Predictive Low-Glucose Alert Users

	Low-glucose alert (N = 163,059)		Predictive low-glucose alert (N = 61,350)
Characteristic	Mean (SD)	Median [25th, 75th]	Mean (SD)	Median [25th, 75th]
Glucose alert use
number of daily set glucose thresholds	1.15 (0.51)	1 [1, 1.01]	1.32 (0.77)	1 [1, 1.05]
glucose alert threshold set (mg/dL)	69.75 (6.73)	70 [68.46, 72.06]	70.6 (7.86)	70 [66.96, 74.67]
time with glucose alert activated (%)	96.49 (14.94)	100 [100, 100]	79.8 (34.21)	100 [68.67, 100]
Time in ranges (%)
Below 54 mg/dL (%)	0.35 (0.47)	0.2 [0.07, 0.44]	0.35 (0.49)	0.19 [0.06, 0.44]
Below 70 mg/dL (%)	1.81 (1.61)	1.38 [0.7, 2.44]	1.78 (1.67)	1.31 [0.63, 2.4]
Within 70–140 mg/dL (%)	47.3 (10.8)	47.3 [40.4, 54.2]	47.1 (11.4)	46.9 [39.8, 54.3]
Within 70–180 mg/dL (%)	71.91 (10.85)	72.83 [65.56, 79.38]	71.78 (11.25)	72.7 [65.19, 79.58]
Above 180 mg/dL (%)	26.29 (11.17)	25.35 [18.58, 32.81]	26.44 (11.62)	25.49 [18.35, 33.3]
Above 250 mg/dL (%)	6.52 (6.18)	4.83 [2.44, 8.63]	6.54 (6.34)	4.81 [2.36, 8.69]
GMI (%)	7.0 (0.43)	6.9 [6.71, 7.2]	7.0 (0.44)	6.9 [6.71, 7.22]

GMI, glucose management indicator.

Figure 1 illustrates the relationship between user-set alert thresholds and time in the different glucose ranges and GMI. Overall time spent in hypoglycemia was low (Fig. 1A and B). Median TB54 and TB70 values were slightly higher at lower alert thresholds compared with higher thresholds; however, across all thresholds, the median, 25th and 75th percentile for TB54 and TB70 remained within recommended targets (<1% and <4%, respectively).

Lower alert thresholds were associated with significantly improved TIR, TITR, and GMI. Median TIR was ∼76% for thresholds <60 mg/dL versus ∼70% for thresholds ≥80 mg/dL (P < 0.0001) (Fig. 1D). Median TITR was ∼51% [44%–58%] for thresholds <60 mg/dL compared with ∼41% [31%–47%] for thresholds ≥80 mg/dL. Median GMI met target of <7% at thresholds <60 mg/dL (6.8% [6.5%–7.0%]) but missed the target at ≥80 mg/dL thresholds (7.1% [6.8%–7.4%]) (Fig. 1C).

The impact of alert thresholds was more pronounced for hyperglycemia metrics (Fig. 1E and F). Alert thresholds above 65 mg/dL were associated with consistently higher TA180 and TA250, just meeting or exceeding recommended targets of <25% and <5%, respectively. Median values for thresholds ≥80 mg/dL were approximately 30% for TA180 and 6% for TA250, compared with ∼22% and ∼3.5%, respectively, for thresholds between <60 mg/dL (P < 0.0001). Similar trends were observed for both LA and PLA users.

Nighttime

Nighttime use patterns of LA and PLA were similar to those observed across the full day. Among the 169,487 individuals in the overall cohort, 96% used LA, 34% used PLA, 34% used both alerts, 62% used only LA, 0.5% used only PLA, and 3.4% used no alerts during nighttime hours. The median percentage time of LA use remained 100% [100%–100%], while PLA use was 100% [60.8%–100%] among users. Median thresholds remained 70 mg/dL [68.5–72 mg/dL] for LA and 70 mg/dL [67–73 mg/dL] for PLA.

Compared with overall daily metrics, nighttime glycemic outcomes were significantly improved for both LA and PLA users (P < 0.0001 for all comparisons), with median nighttime TB54, TB70, TIR, TA180, and TA250 of 0.13%, 0.91%, 80%, 18.6%, and 2.79%, respectively, in LA users, and 0.13%, 0.88%, 80%, 18.5%, and 2.75%, respectively, in PLA users. For all low glucose thresholds, the median time in the several ranges was within the recommended targets (Fig. 2A, B, and D–F), while only for thresholds above 85 mg/dL the median GMI slightly exceeded the threshold of 7% (Fig. 2C).

Younger cohort

Alert usage patterns in the younger cohort were similar to those observed in the overall population. Among 39,597 users aged ≤15 years, 97% used LA, 39% used PLA, 38% used both alerts, 59% used only LA, 0.3% used only PLA, and 2.4% used no alerts.

Figure 3 shows the relationship between alert thresholds and time in glucose ranges for this age group. Overall glycemic control was more variable and slightly worse than in the overall cohort, but similar patterns were observed. The median, 25th and 75th percentile for time in hypoglycemia remained low across thresholds, with values slightly higher at lower alert thresholds but still remained within recommended targets. Lower alert thresholds were associated with higher TIR, whereas alert thresholds above 70 mg/dL were associated with median TIR values at or below the 70% clinical target.

Alert thresholds ≥60 mg/dL were consistently associated with higher time in hyperglycemia, exceeding clinical targets. Median values were 29%–33% for TA180 and 7%–9% for TA250 for thresholds between 80 and 90 mg/dL, compared with 22%–24% and 4%–5%, respectively, for thresholds <60 mg/dL. Median GMI was below 7% for thresholds <65 mg/dL. Similar trends were observed for LA and PLA users, with slightly better glycemic outcomes among users of predictive alerts at lower thresholds.

Achievement of clinical targets

Table 2 shows the proportion of users meeting clinical glycemic targets for the several time in ranges according to the alert threshold category. Across thresholds, a large majority of users met the recommended hypoglycemia targets for TB70 and TB54. Achievement of these targets exceeded 90% of users for thresholds ≥60 mg/dL. In contrast, targets for TIR and hyperglycemia (TA180 and TA250) were more frequently achieved with lower alert thresholds (<60 mg/dL). For users with thresholds below 60 mg/dL, 70% of them met the TIR target, 61% met hyperglycemia targets, and 69% met GMI target. Overall, 43% of users with thresholds below 60 mg/dL met all glycemic targets simultaneously, compared with 30% of users with thresholds above 80 mg/dL.

Table 2.

The Proportion of Users Meeting Clinical Glycemic Targets for Time in Ranges According to Alert Threshold Category

Low-glucose alert threshold (mg/dL)	N	TB54 <1%	TB70 <4%	TIR >70%	TA180 <25%	TA250 <5%	GMI <7%	All targets met
<60	6583	5642 (85.7%)	5390 (81.9%)	4590 (69.7%)	4024 (61.1%)	4025 (61.1%)	4569 (69.4%)	2848 (43.3%)
60 to <70	60,944	56,021 (91.9%)	54,707 (89.8%)	37,434 (61.4%)	30,431 (49.9%)	31,214 (51.2%)	35,562 (58.4%)	23,634 (38.8%)
70 to <80	79,448	74,531 (93.8%)	73,541 (92.6%)	49,087 (61.8%)	39,011 (49.1%)	41,813 (52.6%)	44,960 (56.6%)	32,162 (40.5%)
80 to ≤90	16,084	15,256 (94.9%)	15,168 (94.3%)	7830 (48.7%)	5839 (36.3%)	6964 (43.3%)	6582 (40.9%)	4803 (29.9%)

TA180, time above 180 mg/dL; TA250, time above 250 mg/dL; TB54, time below 54 mg/dL; TB70, time below 70 mg/dL; TIR, time in range.

ROC analysis demonstrated poor discriminatory ability of the low alert thresholds for identifying individuals meeting targets of TB54 <1% and TB70 <4%, with an area under the curve of 0.56 and 0.57, respectively. In addition, there was a weak inverse relationship between the low alert threshold and both TB54 and TB70, as reflected by Spearman correlation coefficients of −0.16 and −0.18, respectively.

Discussion

In this large real-world cohort of MiniMed 780G users across the EMEA region, most individuals enabled LA (96%), whereas fewer used PLA (36%). Only 3% did not activate any low alert; consequently, their low alerts were set at the defaulted level of 54 mg/dL. Alert thresholds were similar between the two alert types, with a median value of 70 mg/dL, which was also the most selected threshold. The lower uptake of predictive alerts may reflect user confidence in the system’s automated hypoglycemia prevention features, while the widespread use of LA suggests that avoidance of hypoglycemia remains a central priority in the management of type 1 diabetes using AID systems.

Time spent in hypoglycemia was low overall. Median and interquartile range values for both TB54 and TB70 remained within recommended clinical targets across all user-set alert thresholds. Although slightly higher hypoglycemia exposure was observed among users with lower alert thresholds, these differences were small and clinically limited. These findings suggest that higher alert thresholds are not necessary to maintain low hypoglycemia exposure when using the MiniMed 780G system, as the algorithm is designed to mitigate hypoglycemia by automatically reducing or suspending insulin delivery when impending hypoglycemia is predicted.

In contrast, user-set alert thresholds had a more pronounced impact on TIR and hyperglycemia metrics. Higher TIR was observed among users with lower alert thresholds, with median TIR reaching approximately 76% at thresholds <60 mg/dL, compared with approximately 67% among users with thresholds ≥80 mg/dL. For these higher thresholds, median TIR approached or fell below the recommended target of 70%. The effect was even more evident for hyperglycemia metrics, where thresholds above 65 mg/dL were consistently associated with higher time above range values. Users with thresholds <60 mg/dL had median TA180 and TA250 values of approximately 22% and 4%, respectively, whereas thresholds ≥80 mg/dL were associated with TA180 and TA250 values exceeding recommended targets. Together, these findings indicate that lower alert thresholds are associated with improved overall glycemic outcomes, characterized by higher TIR and lower time in hyperglycemia while maintaining low time in hypoglycemia.

Nighttime use of glucose alerts and their corresponding thresholds was similar to overall daily use. Glycemic outcomes were significantly improved overnight compared with daytime across all thresholds, with higher TIR and lower exposure to both hypoglycemia and hyperglycemia. Lower alert thresholds maintained low hypoglycemia exposure while further improving TIR and reducing hyperglycemia during nighttime hours. These findings are consistent with previous evidence demonstrating that the MiniMed 780G system provides particularly stable and safe glycemic control during overnight periods.

Users aged ≤15 years demonstrated more variable and overall, less optimal glycemic control compared with the overall cohort, although similar patterns were observed regarding the relationship between alert thresholds and glycemic outcomes. Time spent in hypoglycemia remained low across all thresholds, with median and interquartile range values within recommended targets, suggesting that lower alert thresholds were safe even in this younger population. However, higher alert thresholds were associated with lower TIR and increased hyperglycemia, indicating that such thresholds may not be optimal for younger users.

We have examined whether other factors could have contributed to the observed differences in glycemic outcomes across the LA threshold groups. Specifically, we assessed AID system factors known to be associated with glycemic outcomes. The mean percentage of CGM use, of time in Auto-mode, the total daily dose, the percentage of automated correction bolus insulin, and the number of user-initiated boluses per day were similar across groups. These findings suggest that the observed differences in glycemic outcomes are primarily attributable to differences in LA threshold settings rather than to these AID-related factors. Interestingly, the use of the two recommended optimal settings, active insulin time of 2 h (AIT2h) and glucose target of 100 mg/dL (GT100), was higher among users with lower LA thresholds compared with those using higher thresholds. The mean proportion of time using AIT2h decreased from 63% to 37%, and the use of GT100 decreased from 65% to 25% across the range from <60 to >80 mg/dL LA thresholds. These results indicate that users who set lower alert thresholds are also more likely to use recommended optimal settings and to prioritize tighter glycemic control, whereas those using higher thresholds and alternative settings tend to prioritize avoidance of hypoglycemia. Importantly, time in hypoglycemia remained within recommended clinical targets across all groups.

When evaluating achievement of recommended glycemic targets, hypoglycemia targets were met by most users across all thresholds, with more than 90% of individuals meeting TB70 and TB54 targets when thresholds were ≥60 mg/dL. In contrast, targets for TIR and hyperglycemia were more frequently achieved at lower alert thresholds, with approximately 70% and 61% of users meeting TIR and hyperglycemia targets, respectively, when thresholds <60 mg/dL were used. These findings highlight that selecting alert thresholds based solely on minimizing hypoglycemia is not advisable.

Identifying thresholds that increase the likelihood of achieving all glycemic targets simultaneously is therefore clinically relevant. About 43% of users with alert thresholds <60 mg/dL met all recommended glycemic targets across glucose ranges. The low alert thresholds showed limited effectiveness in identifying individuals meeting TB54 <1% and TB70 <4%, with AUCs of 0.56 and 0.57, respectively. The weak inverse correlations (Spearman’s ρ = −0.16 and −0.18) further highlight the minimal association between the alert thresholds and these target outcomes.

From a clinical perspective, LA thresholds <60 mg/dL may be appropriate for most individuals if there is limited fear of hypoglycemia and improvement in hyperglycemia is desired. Conversely, thresholds between 60 and 70 mg/dL may be preferable for individuals with greater concern about hypoglycemia, accepting slightly less optimal hyperglycemia control.

One limitation of this study relates to the constraints inherent to data collected through CareLink Personal. Several sociodemographic variables, including sex, age, and diabetes type, are self-reported due to privacy regulations, which may introduce misclassification. In addition, age was available only in grouped categories, limiting the granularity of age-specific analyses. Although multiple steps were taken to mitigate bias—including the use of a large dataset, a high proportion of users with CareLink Personal accounts, high consent rates, and automated nightly data uploads—the potential impact of missing or incomplete data on the reported outcomes cannot be fully excluded. Despite these limitations, this study has notable strengths. It leverages a large, geographically diverse real-world population derived from a well-characterized data repository, which enhances the robustness of the analyses. The scale and representativeness of the dataset help reduce selection bias and support the generalizability of the findings across a broad range of clinical settings.

In conclusion, in this large real-world study of MiniMed 780G users, time in hypoglycemia remained low and within recommended clinical targets across all low alert thresholds. A glucose alert threshold <60 mg/dL is generally recommended, as it is associated with increased TIR, reduced hyperglycemia, and sustained low time in hypoglycemia. In some cases, a threshold between 60 and 70 mg/dL may represent a balanced approach for individuals with greater concern about hypoglycemia. Utilizing a low alert threshold may optimize system usability and glycemic outcomes while maintaining safety and minimizing alarm burden across diverse patient populations and care settings.

Authors’ Contributions

J.C. and O.C. conceptualized the study. J.C., M.L., and J.J. prepared and curated the data. J.C. performed the statistical analyses. J.C. and O.C. drafted the article. All authors contributed to the study design, interpretation of the results, and critically reviewed and edited the article. All authors approved the final version for submission. O.C. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the analyses.

Footnotes

Author Disclosure Statement

J.C., M.L., J.J., and O.C. are employees of MiniMed.

Funding Information

This study was funded by MiniMed.

References

1. Ratner

. Hypoglycemia: New Definitions and Regulatory Implications. Diabetes Technol Ther 2018;20(S2):S250–Ss253.

2.Prolonged nocturnal hypoglycemia is common during 12 months of continuous glucose monitoring in children and adults with type 1 diabetes. Diabetes Care 2010;33:1004–1008.

3. Secrest

, Becker

, Kelsey

, et al. Characterizing sudden death and dead-in-bed syndrome in Type 1 diabetes: Analysis from two childhood-onset Type 1 diabetes registries. Diabet Med 2011;28(3):293–300.

4. Abraham

, Arunachalam

, Zhong

, et al. Improved Real-world glycemic control with continuous glucose monitoring system predictive alerts. J Diabetes Sci Technol 2021;15(1):91–97.

5. Giza

, Kotanidou

, Tsinopoulou

, et al. Can Glucose alarm fatigue threaten the absolute clinical benefit of continuous glucose monitoring in optimal glucose management in children and adolescents with type 1 diabetes? A narrative review. Children (Basel) 2025;12(12):1668.

6. Bosi

, Choudhary

, de Valk

, SMILE Study Group. et al. Efficacy and safety of suspend-before-low insulin pump technology in hypoglycaemia-prone adults with type 1 diabetes (SMILE): an open-label randomised controlled trial. Lancet Diabetes Endocrinol 2019;7(6):462–472.

7. Grosman

, Roy

, Lintereur

, et al. A Peek Under the Hood: Explaining the MiniMed™ 780G Algorithm with Meal Detection Technology. Diabetes Technol Ther 2024;26(S3):17–23.

8. Castañeda

, de Galan

, van Kuijk

SMJ

, et al. The interdependence of targets for continuous glucose monitoring outcomes in type 1 diabetes with automated insulin delivery. Diabetes Obes Metab 2024;26(12):5836–5844.

9. Castañeda

, Arrieta

, van den Heuvel

, et al. Time in tight glucose range in type 1 diabetes: Predictive factors and achievable targets in real-world users of the minimed 780G System. Diabetes Care 2024;47(5):790–797.

10.

10. Choudhary

, Arrieta

, van den Heuvel

, et al. Celebrating the Data from 100,000 Real-World Users of the MiniMed™ 780G System in Europe, Middle East, and Africa Collected Over 3 Years: From data to clinical evidence. Diabetes Technol Ther 2024;26(S3):32–37.

11.

11. Tauschmann

, Forlenza

, Hood

, et al. ispad clinical practice consensus guidelines 2022: Diabetes technologies: Glucose monitoring. Pediatr Diabetes 2022;23(8):1390–1405.

12.

12. Lin

, Groat

, Chan

, et al. Alarm settings of continuous glucose monitoring systems and associations to glucose outcomes in type 1 diabetes. J Endocr Soc 2020;4(1):bvz005.

13.

13. Wolpert

. Establishing a continuous glucose monitoring program. J Diabetes Sci Technol 2008;2(2):307–310.

14.

14. Oriot

, Hermans

. Intermittent-scanned continuous glucose monitoring with low glucose alarms decreases hypoglycemia incidence in middle-aged adults with type 1 diabetes in real-life setting. J Diabetes Complications 2023;37(2):108385.

15.

15. González-Vidal

, Rivas-Otero

, Agüeria-Cabal

, et al. Continuous glucose monitoring alarms in adults with type 1 diabetes: User characteristics and the impact of hypoglycemia and hyperglycemia alarm thresholds on glycemic control. Diabetes Technol Ther 2024;26(5):313–323.