In a recent study published in the journal Nature Medicineresearchers examined fasting glucose (FG) variability in nondiabetic adults using continuous glucose monitoring (CGM), assessing its impact on diabetes classification and its association with clinical measures.
Study: Continuous glucose monitoring and intraindividual variability in fasting glucose. Image credit: Suriyawut Suriya / Shutterstock
Record
The global increase in prediabetes and diabetes poses significant health risks and economic burdens. Diagnosis is mainly based on elevated fasting plasma glucose (PFG), glycated hemoglobin (HbA1c) levels, oral glucose tolerance test (OGTT) or random plasma glucose in symptomatic individuals. However, OGTT is often bypassed due to its cost and inconvenience, leaving PFG and HbA1c as the main diagnostic tools, especially for asymptomatic cases. Despite its diagnostic importance, the day-to-day variability of PFG in non-diabetic individuals remains unexplored, potentially leading to misdiagnosis. CGM devices, which measure interstitial glucose levels, offer improved accuracy over time and are now used independently or in hybrid closed-loop systems for insulin dosing. Further research is needed to develop CGM-based diagnostic criteria that accurately reflect the interindividual variability of FG levels and their clinical implications.
About the study
The present study analyzed data from the 10K study, focusing on people aged 40 to 70 years. At baseline, various measures including lifestyle, dietary habits, vital signs and medical history were collected alongside specific tests such as blood tests, electrocardiography and CGM using the FreeStyle Libre Pro Flash system for two weeks. This study included participants without a self-reported diagnosis of type 2 diabetes or related conditions who also participated in active meal recording along with their CGM data. Exclusion criteria were strict, including abnormal CGM readings and inadequate meal recording.
The research placed particular emphasis on FG measurements during the morning hours, using CGM data to observe intra-individual variability and its potential impact on diabetes diagnosis. The methodology ensured a realistic simulation of fasting conditions, relying on at least 8 hours of no caloric intake before the measurement windows and strict meal recording criteria. A total of 8,315 subjects with 59,565 morning fasting windows were analyzed for FG variability and its association with various clinical measures, including anthropometry, vital signs, and sleep monitoring, among others. Sleep monitoring used the Food and Drug Association (FDA)-the WatchPAT-300 device was approved, while detailed retinal imaging and other health metrics were carefully analyzed for associations with FG variability. In addition, the study applied statistical analyzes to investigate the relationship between FG variability and clinical measurements, taking age and gender into account.
Study results
By analyzing FG readings from 8,315 people in 59,565 morning windows, the researchers delved into the details of FG variability and its implications for diabetes classification. Study participants, on average 51.3 years old, had a mean body mass index (BMI) of 25.92 ± 4.07 kg m−2. Data collection was rigorous, with morning FG measurements performed between 06:00 and 09:00, after a pre-specified minimum of 8 hours of fasting, although the actual mean duration of fasting was over 10 hours. Notably, fasting duration did not show a significant correlation with FG values.
Study methodology was thorough in calculating FG for each individual, ensuring valid morning windows through strict criteria, including active meal recording. This large-scale examination revealed a mean FG value of 96.2 mg dl−1, which was observed to increase slightly with age, indicating a gradual increase in glucose levels over time. The analysis also highlighted significant day-to-day variability in FG measurements across subjects, a finding that highlights the complex nature of glucose metabolism and its sensitivity to various factors.
When assessing the potential for misclassification of diabetes and pre-diabetes based on FG levels, the research found considerable variability. A notable proportion of participants experienced changes in their glycemic status classification throughout the study, emphasizing the limitations of relying on a single FG measurement to diagnose diabetes. This variability, combined with the narrow range defining normal and diabetic FG levels, suggests the need for refined diagnostic criteria to better accommodate individual variations in glucose readings.
The study also examined the clinical associations of FG variability with various health markers, including body composition, blood pressure, and liver function. Interestingly, FG variability showed significant correlations with several clinical measures, underscoring its potential as a marker for metabolic health. In particular, associations with body composition and daily caloric intake suggest that FG variability may reflect broader metabolic processes beyond glucose regulation alone.
conclusions
In summary, this research analyzed FG data from 8,315 nondiabetic subjects using CGM, revealing significant FG variability that calls into question the reliability of current PFG-based diabetes diagnostic criteria. Initial classifications showed that most participants had normal FG levels, but further measurements suggested a significant shift towards prediabetes, highlighting the risk of misclassification. The study showed that increasing the number of FG tests could significantly reduce misdiagnosis. In addition, it found significant correlations between FG levels and various clinical measures within normal glucose limits, suggesting the need for a better approach to diabetes diagnosis that takes into account the variability and dynamic nature of FG levels.
