Continuous Glucose Monitoring Profiles in Healthy Nondiabetic Participants: A Multicenter Prospective Study

In the paper

“Our findings support the recent consensus conference recommendation (12, 13) that sensor glucose values ,54 mg/dL represent a more meaningful cut-off than 70 mg/dL to define clinically important hypoglycemia because CGM glucose levels below 54 mg/dL were uncommon, whereas most participants had some glucose levels of 55 to 69 mg/dL.”

What the sources say

• Supported

  The hypoglycemia recommendations classify sensor glucose below 54 mg/dL as clinically significant and 54–70 mg/dL as an alert range. They also recommend giving values below 54 mg/dL more weight in CGM outcome reports. The citing paper’s observation about how often participants entered each range is its own finding, not part of the recommendation attributed to this source. (LLM, conf=0.94)

  cites [12] Danne, 2017, International consensus on use of continuous glucose monitoring, Diabetes Care source
• Partly supported

  In Additional Recommendations and Limitations (page unavailable), the consensus report suggests reporting CGM excursions below either 70 or 55 mg/dL as possible measures of hypoglycemia. It does not recommend a cutoff below 54 mg/dL or say that cutoff is more clinically meaningful than 70 mg/dL. (LLM, conf=0.85)

  cites [13] Maahs, 2016, Outcome measures for artificial pancreas clinical trials: a consensus report, Diabetes Care sourceExperimental — accuracy not yet measured

In the paper

“Most studies of CGM profiles in healthy, nondiabetic individuals have had small sample sizes or used early-generation CGM systems, which were less accurate than current devices (3)(4)(5)(6)(7).”

What the sources say

• Partly supported

  The Discussion describes earlier CGM studies of 32 and 15 people, supporting the small-sample part of the claim; Methods reports that this study used CGM in 80 nondiabetic participants. Neither section says the systems were early-generation or compares their accuracy with current devices. (LLM, conf=0.84)

  cites [3] Borg, 2010, Real-life glycaemic profiles in non-diabetic individuals with low fasting glucose and normal HbA1c: the A1C-Derived Average Glucose (ADAG) study, Diabetologia sourceExperimental — accuracy not yet measured
• Partly supported

  The Methods report CGM profiles from 24 healthy volunteers using SCGM1 devices, supporting this paper as an example of a small study using an older system. The paper reports the SCGM1’s measurement error but does not compare its accuracy with current devices or establish that most such studies share these limitations. (LLM, conf=0.84)

  cites [4] Freckmann, 2007, Continuous glucose profiles in healthy subjects under everyday life conditions and after different meals, J Diabetes Sci Technol sourceExperimental — accuracy not yet measured
• Partly supported

  The Abstract’s Research Design and Methods describes a CGM study of 74 healthy participants. Research Design and Methods identifies the devices used and reports missing or implausible readings from one sensor, noting that a commercially available version had newer software. The paper does not establish that most such studies had small samples or that earlier devices were less accurate than current devices. (LLM, conf=0.83)

  cites [5] Fox, 2010, Variation of interstitial glucose measurements assessed by continuous glucose monitors in healthy, nondiabetic individuals, Diabetes Care sourceExperimental — accuracy not yet measured

In the paper

“As the assessment of outcomes of treatment of T1D and type 2 diabetes in clinical practice and research trials moves beyond the narrow focus of HbA1c levels to include the results of sensor glucose profiles (14), there is an increased need for a repository of”

What the sources say

• Supported

  The citation is attached to the clause about assessing diabetes outcomes beyond HbA1c using sensor glucose profiles, not to the subsequent need for a normative-data repository. The Introduction says CGM profiles provide information missed by HbA1c, and the Conclusions say a patient’s CGM glucose profile has considerable value for diabetes management. (LLM, conf=0.88)

  cites [14] Beck, 2017, The fallacy of average: how using HbA1c alone to assess glycemic control can be misleading, Diabetes Care source

In the paper

“This is similar to what has been observed in the T1D population and highlights the limitation of HbA1c to reflect true glycemic status for some individuals (14).”

What the sources say

• Supported

  The Introduction states that a wide range of mean glucose concentrations can accompany a given HbA1c and that this has been demonstrated in people with type 1 diabetes. It also explains that HbA1c alone may therefore misrepresent an individual’s glycemic control. Page numbers were not supplied for these passages. (LLM, conf=0.91)

  cites [14] Beck, 2017, The fallacy of average: how using HbA1c alone to assess glycemic control can be misleading, Diabetes Care source

In the paper

“The current consensus conferences have somewhat arbitrarily set time in ranges for CGM sensor glucose levels for patients with T1D, with time in target identified as being between 70 to 180 mg/dL (12,13).”

What the sources say

• Supported

  The ATTD consensus table (p. 1636) lists 70–180 mg/dL as the default CGM time-in-target range, supporting the attributed range for patients with type 1 diabetes. The paper does not describe that cutoff as “somewhat arbitrary”; that appears to be the citing authors’ characterization. (LLM, conf=0.90)

  cites [12] Danne, 2017, International consensus on use of continuous glucose monitoring, Diabetes Care source
• Partly supported

  The Objective and Rationale section advocates standardized CGM glucose metrics for artificial-pancreas studies and says they are broadly applicable to type 1 diabetes studies. Additional Recommendations and Limitations mentions arbitrary glycemic ranges, but the retrieved passages do not identify 70–180 mg/dL as the target range or say that consensus conferences set that range somewhat arbitrarily. Page numbers were not available in our copy. (LLM, conf=0.80)

  cites [13] Maahs, 2016, Outcome measures for artificial pancreas clinical trials: a consensus report, Diabetes Care sourceExperimental — accuracy not yet measured

In the paper

“A total of 153 participants (age 7 to 80 years) were included in the analyses.”

• Also stated in: § Results · plain

  “Therefore, 153 participants were included in the analyses, ranging from 7 to 80 years (mean, 31 years).”

In the paper

“Median sensor glucose between 70 and 140 mg/dL during the daytime and nighttime was 96% (IQR, 92 to 97%)”

In the paper

“Sensor glucose values .180mg/dL (10.0 mmol/L) and ,54 mg/dL (3.0 mmol/L) were uncommon.”

In the paper

“In a sensitivity analysis after removing participants with first-degree family members with T1D (n 5 101), mean CGM glucose, time spent in the 70 to 140 mg/dL range, and glycemic variability as measured by CV were not different from the main analysis (data not shown).”

In the paper

“Mean average glucose was 98 to 99 mg/dL (5.4 to 5.5 mmol/L) for all age groups except those over 60 years, in whom mean average glucose was 104 mg/dL (5.8 mmol/L).”

• Also stated in: § Results · plain

  “The mean average glucose level was 98 to 99 mg/dL (5.4 to 5.5 mmol/L) for all age groups except those $60 years old in whom mean average glucose a The underweight, normal weight, and overweight BMI categories for participants aged $18 years are ,18.5, 18.5 to ,24.9, and $24.9, respectively.”

In the paper

“Similarly, in a sensitivity analysis after removing implausible data, mean glucose, time spent in the 70 to 140 mg/dL range, and CV were not different from the main analysis”

In the paper

“mean glycemic variability measured by CV during the daytime and nighttime was 17% 6 3%”

In the paper

“Mean CGM use by the participants was 192 hours (8 days)”

In the paper

“with participants 6 to 11 years old having a median of 90% of their readings in this range, as compared with only 81% in the cohort over age 60 years.”

In the paper

“There are two possible factors that may have contributed to this weaker-than-expected correlation: (1) HbA1c was obtained prior to collection of CGM glucose data and ( 2) there are at most 10 days of CGM data (16).”

What the sources say

• ?Not checked — no full text of the sourcecites [16] Riddlesworth, 2018, Optimal sampling duration for continuous glucose monitoring to determine long-term glycemic control, Diabetes Technol Ther source

In the paper

“Understanding CGM metrics in healthy participants without diabetes is needed to serve as a benchmark for research studies to define impaired glycemic status and for future technologies aiming to help individuals achieve near-normal glycemic profiles.”

In the paper

“Strengths of our study include the large sample size crossing the lifespan, multiple centers conducting the study, use of the latest generation technology, and up to 10 days of CGM wear time.”

In the paper

“Although our study fills the current gap in the data”

In the paper

“The reported mean absolute relative difference (MARD) for the Dexcom G6 in accuracy studies comparing sensor glucose values with reference blood glucose values is 9.0% (9).”

What the sources say

• ?Not checked — no full text of the sourcecites [9] Shah, 2018, Performance of a factory-calibrated real-time continuous glucose monitoring system utilizing an automated sensor applicator, Diabetes Technol Ther source

In the paper

“Although pressure on the sensor during sleep may have resulted in inaccuracy of some CGM readings”

In the paper

“CGM profiles also provide the best window available to date to evaluate the frequency of biochemical hypoglycemia in clinical trials of new drugs and devices for the treatment of diabetes.”

In the paper

“However, HbA1c does not capture clinically relevant hyperglycemic and hypoglycemic patterns, which limits its utility in personalizing insulin-dosing decisions.”

In the paper

“Although data on glycemic profiles of healthy, nondiabetic individuals exist for older sensors, assessment of glycemic metrics with new-generation CGM devices is lacking.”

In the paper

“This is due to the fact that HbA1c is a reflection of overall glycemic control from the prior 60-90 days, as well as clear evidence from the DCCT establishing it as a surrogate marker for diabetes microvascular complications risk (1).”

What the sources say

• ?Not checked — no full text of the sourcecites [1] Nathan, 1993, The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus, N Engl J Med source

In the paper

“Use of continuous glucose monitoring (CGM) is increasing for insulin-requiring patients with diabetes.”

In the paper

“Participants were healthy, nondiabetic children and adults, primarily recruited from family, friends, and neighbors of patients seen in the diabetes clinics.”

In the paper

“Moreover, CGM metrics are being used to examine and compare the relative effectiveness and safety of new therapeutic agents and modern technologies in clinical diabetes research.”

In the paper

“Continuous glucose monitoring (CGM) provides information on trends for hypoglycemia, hyperglycemia, and glycemic variability that helps patients and providers optimize glycemic control without increasing the risk for hypoglycemia.”

In the paper

“Study participants provided written informed consent prior to study participation.”

In the paper

“Only participants with $72 total hours of CGM data (with at least 24 hours overnight) were included in the analysis.”

In the paper

“Only participants with no islet autoantibodies and a central laboratory HbA1c level ,5.7% were included in the analysis.”

In the paper

“The study was conducted at 12 diabetes centers within the T1D Exchange Clinic Network after approval by institutional review boards (8).”

What the sources say

• ?Not checked — no full text of the sourcecites [8] Beck, 2012, The T1D Exchange clinic registry, J Clin Endocrinol Metab source