A 13-year-old girl with childhood-onset obesity was diagnosed with type 1 diabetes. At the time of the diagnosis, her hemoglobin A1c (HbA1c) level was 12.3%, anti-GAD and anti–islet cell antibodies were positive, she had not developed diabetes ketoacidosis (DKA), and her BMI was 21.8 kg/m2 (85th percentile). Treatment with an insulin pump (Medtronic MiniMed 640G) and a continuous glucose monitoring system was initiated. Her follow-up was characterized by HbA1c levels between 7.8% and 8.5%, along with marked weight gain.
At age 14.5 years, she was hospitalized with mild DKA, attributed to pump malfunction, and her BMI was 28.9 kg/m2 (96th percentile). Several weeks later, her glycemic metrics revealed a time in range (TIR) of 34%, time above range of 65.6%, time below range of 0.4%, and HbA1c level of 8.7%. The characteristics of a typical day are depicted in Fig. 1A. Insulin omission for weight loss was suspected.
Daily summary reports of continuous glucose monitor, insulin delivery and events from three different periods. A: A typical day, showing elevated glucose levels, unreported carbohydrate (carb) intake, and late and missed insulin doses for both meals and correction. B: Improved glycemic metrics, despite missed insulin doses after the switch to MiniMed 780G. C: Virtually perfect TIR, secondary to calibration based on false glucose reading. TDD, total daily dose; U, units.
Daily summary reports of continuous glucose monitor, insulin delivery and events from three different periods. A: A typical day, showing elevated glucose levels, unreported carbohydrate (carb) intake, and late and missed insulin doses for both meals and correction. B: Improved glycemic metrics, despite missed insulin doses after the switch to MiniMed 780G. C: Virtually perfect TIR, secondary to calibration based on false glucose reading. TDD, total daily dose; U, units.
To mitigate intentionally reduced or missed insulin administration and minimize high blood glucose levels, treatment with an advanced hybrid closed-loop (AHCL) system—the Medtronic MiniMed 780G—was initiated. The 780G system worked with a user-calibrated Guardian 3 sensor, which had to be calibrated twice daily, and automatically adjusted basal insulin delivery along with automatic correction boluses based on continuous glucose monitor readings. Her glucose metrics improved markedly within a few weeks. The TIR was 60%, time above range was 38%, time below range was 2%, and glucose management indicator (GMI) was 7.6% (Fig. 1B). In view of this improvement, 3 months after switching to 780G, her family opted for a virtual rather than face-to-face visit. The downloaded data were promising, with TIR of up to 97% (Fig. 1C).
At age 15 years, she was admitted to the hospital with impaired consciousness. In the evening before her admission, she felt sick, her glucose levels displayed by the Guardian 3 ranged between 70 and 180 mg/dL, and the inserted glucose level for calibration was 167 mg/dL. However, in the emergency room, her glucose level was 694 mg/dL, pH was 6.99, HCO3 was 6.9 nmol/L (23–27), and HbA1c was 9.7%. At the same time, the GMI in the CareLink system was 7.0% and the average blood glucose ± SD was 156 ± 40 mg/dL. SmartGuard use was 97%, and sensor wear was 86%.
The parents expressed distrust in the Guardian 3. However, an examination of her glucometer revealed that she had not tested her glucose level for a month. For calibration, she entered fake numbers. The AHCL algorithm had adjusted levels according to these falsified calibrations; therefore, it appeared she was achieving consensus targets, while her actual glucose values were likely much higher.
Deliberate insulin restriction or omission for weight loss is increasingly recognized as a major problem in adolescents with type 1 diabetes. Prominent risk factors are female gender, increased weight concerns, and body dissatisfaction (1). According to a nationwide study (2), 27.7% of female adolescents and 8.6% of male adolescents who received intensified insulin treatment had disturbed eating behavior, and one-third occasionally omit their insulin dose entirely after overeating. Insulin omission results in suboptimal glycemic targets, is the primary cause of DKA (3), and is associated with dramatically increased morbidity and mortality rates (4).
The new technology of AHCL systems has proven beneficial for adolescents with suboptimal glycemic metrics (5). Capable of correcting postprandial glucose excursions by increasing insulin delivery, it is, therefore, a possible solution for adolescents with insulin omission. By facilitating insulin delivery and improving glycemic targets, however, the AHCL system may cause excessive weight gain.
With the use of AHCL now widespread, we want to increase awareness of strategies adolescents use to evade insulin administration. One option is disconnecting from Auto Mode prior to meals and skipping insulin administration. Since glucose levels will rise, the algorithm will eventually deliver insulin when reconnected. A more sophisticated method is to calibrate the system with fictitious values. Even in the new Guardian 4, glucose values can be calibrated. In our case, not only was no insulin administered, but the seemingly normal range of glucose levels led to reduced parental supervision and the belief that an in-person clinic visit was unnecessary. A virtual visit that relied on deceptive data and GMI could not uncover the problem. In retrospect, however, a warning sign of false calibration was the perfect match between SmartGuard and blood glucose values.
In summary, we present a case of a female adolescent using the 780G system and fabricating values to calibrate glucose levels as a way of omitting insulin for weight loss. With the increased use across a broad population, caregivers should be alert to potential ways of circumventing this advanced technology.
Article Information
Duality of Interest. O.P.-H. participated in a study conducted by Medtronic on the 670 insulin pump. No other potential conflicts of interest relevant to this article were reported.
Author Contributions. O.P.-H. wrote the manuscript. N.L. and R.F.-H. reviewed the manuscript and contributed to the discussion. O.P.-H. 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 data analysis.