Continuous Glucose Monitors Overestimate Blood Sugar Responses to Food
Study reveals CGMs consistently overestimate glycemic responses, potentially misclassifying food glycemic index ratings.
Summary
Researchers compared continuous glucose monitors (CGMs) to the gold standard fingerstick blood tests across seven different food challenges in 15 healthy adults. CGMs consistently overestimated both fasting and post-meal glucose levels by about 0.9 mmol/L. The bias varied between individuals and different foods, sometimes changing a food's glycemic index classification from low to moderate. CGMs overestimated time spent above 7.8 mmol/L by approximately 4-fold. While adjusting for baseline differences reduced this overestimate to 2-fold, significant inaccuracies remained. This challenges the growing use of CGMs for dietary guidance in healthy individuals.
Detailed Summary
As continuous glucose monitors (CGMs) gain popularity among health-conscious consumers for tracking food responses, new research reveals significant accuracy concerns that could mislead dietary decisions. A rigorous crossover study compared CGM readings to the gold standard fingerstick blood glucose measurements across multiple food challenges.
Fifteen healthy adults completed seven laboratory visits, consuming different carbohydrate sources including pure glucose, whole fruits, blended fruits, and commercial fruit smoothies with various modifications. Researchers measured glucose responses every 15 minutes for two hours using both CGMs and capillary blood sampling.
The results showed CGMs consistently overestimated glucose levels by an average of 0.9 mmol/L for both fasting and post-meal readings. This bias wasn't uniform—it varied significantly between individuals and different food types. Most concerning, the overestimation sometimes changed a food's glycemic index classification, with one fruit smoothie shifting from low glycemic index (53) when measured by blood to moderate glycemic index (69) when measured by CGM.
CGMs also dramatically overestimated time spent in elevated glucose ranges, showing participants spent about 4 times longer above 7.8 mmol/L than blood measurements indicated. Even after adjusting for individual baseline differences, CGMs still overestimated this metric by roughly 2-fold.
These findings have important implications as CGMs become mainstream tools for metabolic health optimization. The technology's tendency to overestimate glucose responses could lead people to unnecessarily restrict foods or make dietary changes based on inaccurate data. For researchers studying food effects on metabolism, the study reinforces that capillary blood sampling remains the most reliable method for accurate glycemic response measurement.
Key Findings
- CGMs overestimated glucose levels by 0.9 mmol/L compared to blood tests
- Bias varied between individuals and food types, affecting glycemic index classification
- Time above 7.8 mmol/L was overestimated 4-fold by CGMs
- Adjustment for baseline differences only partially corrected the overestimation
- Fruit smoothie glycemic index changed from low (53) to moderate (69) with CGM
Methodology
Randomized crossover trial with 15 healthy adults completing 7 laboratory visits, comparing CGM (Abbott Freestyle Libre 2) to capillary blood glucose measurements across different carbohydrate challenges including glucose, whole fruits, and fruit smoothies.
Study Limitations
Study limited to healthy adults and specific food types; results may not apply to people with diabetes or other metabolic conditions. Only one CGM brand tested, and individual variation in accuracy was substantial.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.